- Physical Review Journals
All JournalsPhysics Magazine
Access Provided by Xinjiang University

Physical Review Fluids

Export citation

Export citation

Choose format for download:

Download Citation
  • Access by Xinjiang University

Online learning of subgrid-scale models for quasigeostrophic turbulence in planetary interiors

Hugo Frezat*, Thomas Gastine, and Alexandre Fournier

  • *Contact author: hugo.frezat@gmail.com
PDF

Phys. Rev. Fluids 11, 104904 – Published 14 October, 2026

DOI: https://doi.org/10.1103/lcd6-p52d

Abstract

Machine learning approaches to subgrid-scale (SGS) modeling are now well established in atmospheric and oceanic applications. Among these, online end-to-end learning, where the differentiable solver participates in the training, has shown particular promise. Yet existing studies are largely restricted to idealized periodic domains, with no mechanical boundaries, precluding them from addressing the dynamics of bounded rotating flows relevant to planetary interiors. Here we consider two-dimensional quasigeostrophic turbulence in a rapidly rotating annular bounded domain. We examine three configurations varying the geometry of the container and the rotation rate. The system exhibits key features such as zonal jets, Rossby waves, and, in the spherical shell geometry, a slow quasiperiodic inward drift of the jets. The spectral properties of the zonal and nonzonal flow can be understood in the framework of zonostrophic turbulence theory. We develop a differentiable solver for this system, which allows us to train SGS models online, over a time span of one turnover time, using coarse-grained data from direct numerical simulations. In all cases a SGS model trained on a single turnover time accurately reproduces global integrated diagnostics, energy spectra as well as long-term dynamical behaviours, such as jet migration, occurring on timescales which exceed the training window by one order of magnitude. The online-trained model further outperforms classical hyperdiffusivity and Leith closure schemes, for which reducing the radial resolution is impractical. The resulting speed-up paves the way to further investigations of long-term geophysical fluid processes beyond the reach of direct numerical simulations.

View figure in article

Physics Subject Headings (PhySH)

Article Text

References (106)

  1. M. Landeau, A. Fournier, H.-C. Nataf, D. Cébron, and N. Schaeffer, Sustaining Earth's magnetic dynamo, Nat. Rev. Earth Environ. 3, 255 (2022).
  2. C. C. Finlay, N. Gillet, J. Aubert, P. W. Livermore, and D. Jault, Gyres, jets and waves in the Earth's core, Nat. Rev. Earth Environ. 4, 377 (2023).
  3. M. Pais and D. Jault, Quasi-geostrophic flows responsible for the secular variation of the Earth's magnetic field, Geophys. J. Int. 173, 421 (2008).
  4. G. A. Glatzmaiers and P. H. Roberts, A three-dimensional self-consistent computer simulation of a geomagnetic field reversal, Nature (London) 377, 203 (1995).
  5. J. Wicht and D. G. Meduri, A Gaussian model for simulated geomagnetic field reversals, Phys. Earth Planet. Inter. 259, 45 (2016).
  6. J. Smagorinsky, General circulation experiments with the primitive equations: I. The basic experiment, Mon. Weather Rev. 91, 99 (1963).
  7. M. Maltrud and G. Vallis, Energy spectra and coherent structures in forced two-dimensional and beta-plane turbulence, J. Fluid Mech. 228, 321 (1991).
  8. Q. Chen and G. A. Glatzmaier, Large eddy simulations of two-dimensional turbulent convection in a density-stratified fluid, Geophys. Astrophys. Fluid Dyn. 99, 355 (2005).
  9. W. Kuang and J. Bloxham, Numerical modeling of magnetohydrodynamic convection in a rapidly rotating spherical shell: Weak and strong field dynamo action, J. Comput. Phys. 153, 51 (1999).
  10. H.-C. Nataf and N. Schaeffer, Turbulence in the core, in Treatise on Geophysics, edited by G. Schubert, 2nd ed. (Elsevier, Amsterdam, 2015), Vol. 8, pp. 161–181.
  11. J. Aubert, T. Gastine, and A. Fournier, Spherical convective dynamos in the rapidly rotating asymptotic regime, J. Fluid Mech. 813, 558 (2017).
  12. M. Germano, A proposal for a redefinition of the turbulent stresses in the filtered Navier-Stokes equations, Phys. Fluids 29, 2323 (1986).
  13. A. Leonard, Energy cascade in large-eddy simulations of turbulent fluid flows, in Advances in Geophysics, edited by F. N. Frenkiel and R. E. Munn (Academic Press, New York, 1975), Vol. 18, pp. 237–248.
  14. J. Bardina, J. H. Ferziger, and W. C. Reynolds, Improved subgrid-scale models for large-eddy simulation, in AlAA 13th Fluid and Plasma Dynamics Conference (American Institute of Aeronautics and Astronautics, Snowmass, CO, USA, 1980), p. 1357.
  15. B. A. Buffett, A comparison of subgrid-scale models for large-eddy simulations of convection in the Earth's core, Geophys. J. Int. 153, 753 (2003).
  16. H. Matsui and B. A. Buffett, Sub-grid scale model for convection-driven dynamos in a rotating plane layer, Phys. Earth Planet. Inter. 153, 108 (2005).
  17. H. Matsui and B. A. Buffett, Commutation error correction for large eddy simulations of convection driven dynamos, Geophys. Astrophys. Fluid Dyn. 101, 429 (2007).
  18. H. Matsui and B. A. Buffett, A dynamic scale-similarity model for dynamo simulations in a rotating plane layer, Geophys. Astrophys. Fluid Dyn. 101, 451 (2007).
  19. H. Matsui and B. A. Buffett, Large-eddy simulations of convection-driven dynamos using a dynamic scale-similarity model, Geophys. Astrophys. Fluid Dyn. 106, 250 (2012).
  20. S. L. Brunton, B. R. Noack, and P. Koumoutsakos, Machine learning for fluid mechanics, Annu. Rev. Fluid Mech. 52, 477 (2020).
  21. R. Vinuesa and S. L. Brunton, Enhancing computational fluid dynamics with machine learning, Nat. Comput. Sci. 2, 358 (2022).
  22. J. Yuval and P. A. O'Gorman, Stable machine-learning parameterization of subgrid processes for climate modeling at a range of resolutions, Nat. Commun. 11, 3295 (2020).
  23. T. Arcomano, I. Szunyogh, A. Wikner, B. R. Hunt, and E. Ott, A hybrid atmospheric model incorporating machine learning can capture dynamical processes not captured by its physics-based component, Geophys. Res. Lett. 50, e2022GL102649 (2023).
  24. T. Bolton and L. Zanna, Applications of deep learning to ocean data inference and subgrid parameterization, J. Adv. Model. Earth Syst. 11, 376 (2019).
  25. T. S. Finn, C. Durand, A. Farchi, M. Bocquet, Y. Chen, A. Carrassi, and V. Dansereau, Deep learning subgrid-scale parametrisations for short-term forecasting of sea-ice dynamics with a Maxwell elasto-brittle rheology, The Cryosphere 17, 2965 (2023).
  26. S. Rasp, M. S. Pritchard, and P. Gentine, Deep learning to represent subgrid processes in climate models, Proc. Natl. Acad. Sci. USA 115, 9684 (2018).
  27. P. D. Dueben and P. Bauer, Challenges and design choices for global weather and climate models based on machine learning, Geosci. Model Dev. 11, 3999 (2018).
  28. J. Yuval, P. A. O'Gorman, and C. N. Hill, Use of neural networks for stable, accurate and physically consistent parameterization of subgrid atmospheric processes with good performance at reduced precision, Geophys. Res. Lett. 48, e2020GL091363 (2021).
  29. M. Gelbrecht, A. White, S. Bathiany, and N. Boers, Differentiable programming for Earth system modeling, Geosci. Model Dev. 16, 3123 (2023).
  30. S. Rasp, Coupled online learning as a way to tackle instabilities and biases in neural network parameterizations: General algorithms and Lorenz 96 case study (v1. 0), Geosci. Model Dev. 13, 2185 (2020).
  31. D. Kochkov, J. A. Smith, A. Alieva, Q. Wang, M. P. Brenner, and S. Hoyer, Machine learning–accelerated computational fluid dynamics, Proc. Natl. Acad. Sci. USA 118, e2101784118 (2021).
  32. H. Frezat, J. Le Sommer, R. Fablet, G. Balarac, and R. Lguensat, A posteriori learning for quasi-geostrophic turbulence parametrization, J. Adv. Model. Earth Syst. 14, e2022MS003124 (2022).
  33. F. E. Yan, H. Frezat, J. Le Sommer, J. Mak, and K. Otness, Adjoint-based online learning of two-layer quasi-geostrophic baroclinic turbulence, J. Adv. Model. Earth Syst. 17, e2024MS004857 (2025).
  34. D. Kochkov, J. Yuval, I. Langmore, P. Norgaard, J. Smith, G. Mooers, M. Klöwer, J. Lottes, S. Rasp, P. Düben, S. Hatfield, P. Battaglia, A. Sanchez-Gonzalez, M. Willson, M. P. Brenner, and S. Hoyer, Neural general circulation models for weather and climate, Nature (London) 632, 1060 (2024).
  35. D. Lemasquerier, B. Favier, and M. Le Bars, Zonal jets experiments in the gas giants' zonostrophic regime, Icarus 390, 115292 (2023).
  36. J. Rotvig, Multiple zonal jets and drifting: Thermal convection in a rapidly rotating spherical shell compared to a quasigeostrophic model, Phys. Rev. E 76, 046306 (2007).
  37. R. Chemke and Y. Kaspi, Poleward migration of eddy-driven jets, J. Adv. Model. Earth Syst. 7, 1457 (2015).
  38. T. L. Lonner, A. Aggarwal, and J. M. Aurnou, Planetary core-style rotating convective flows in paraboloidal laboratory experiments, J. Geophys. Res.: Planets 127, e2022JE007356 (2022).
  39. L. Cope, The dynamics of geophysical and astrophysical turbulence, Ph.D. thesis, University of Cambridge, 2021.
  40. S. Cabanes, J. Aurnou, B. Favier, and M. Le Bars, A laboratory model for deep-seated jets on the gas giants, Nat. Phys. 13, 387 (2017).
  41. H.-Y. Zhu, T.-Y. Pei, J.-H. Xie, and K.-Q. Xia, Zonal jets over small-scale topography: Jet properties and energy transfers, J. Fluid Mech. 1020, A35 (2025).
  42. J. Aubert, N. Gillet, and P. Cardin, Quasigeostrophic models of convection in rotating spherical shells, Geochem. Geophys. Geosyst. 4, 2002GC000456 (2003).
  43. C. Guervilly and P. Cardin, Multiple zonal jets and convective heat transport barriers in a quasi-geostrophic model of planetary cores, Geophys. J. Int. 211, 455 (2017).
  44. N. Gillet and C. Jones, The quasi-geostrophic model for rapidly rotating spherical convection outside the tangent cylinder, J. Fluid Mech. 554, 343 (2006).
  45. T. Gastine, pizza: An open-source pseudo-spectral code for spherical quasi-geostrophic convection, Geophys. J. Int. 217, 1558 (2019).
  46. T. Gastine, GitHub, 2026, https://github.com/magic-sph/pizza.
  47. E. Plaut and F. H. Busse, Low-Prandtl-number convection in a rotating cylindrical annulus, J. Fluid Mech. 464, 345 (2002).
  48. H. P. Greenspan, The Theory of Rotating Fluids (Cambridge University Press, Cambridge, 1968).
  49. N. Schaeffer and P. Cardin, Quasigeostrophic model of the instabilities of the Stewartson layer in flat and depth-varying containers, Phys. Fluids 17, 104111 (2005).
  50. S. Boscarino, L. Pareschi, and G. Russo, Implicit-explicit Runge–Kutta schemes for hyperbolic systems and kinetic equations in the diffusion limit, SIAM J. Sci. Comput. 35, A22 (2013).
  51. V. Gopinath, A. Fournier, and T. Gastine, An assessment of implicit-explicit time integrators for the pseudo-spectral approximation of Boussinesq thermal convection in an annulus, J. Comput. Phys. 460, 110965 (2022).
  52. R. Maulik, O. San, A. Rasheed, and P. Vedula, Subgrid modelling for two-dimensional turbulence using neural networks, J. Fluid Mech. 858, 122 (2019).
  53. Y. Guan, A. Chattopadhyay, A. Subel, and P. Hassanzadeh, Stable a posteriori LES of 2D turbulence using convolutional neural networks: Backscattering analysis and generalization to higher Re via transfer learning, J. Comput. Phys. 458, 111090 (2022).
  54. J. Sirignano, J. F. MacArt, and J. B. Freund, DPM: A deep learning PDE augmentation method with application to large-eddy simulation, J. Comput. Phys. 423, 109811 (2020).
  55. H. Frezat, G. Balarac, J. Le Sommer, R. Fablet, and R. Lguensat, Physical invariance in neural networks for subgrid-scale scalar flux modeling, Phys. Rev. Fluids 6, 024607 (2021).
  56. Z. Zhou, G. He, S. Wang, and G. Jin, Subgrid-scale model for large-eddy simulation of isotropic turbulent flows using an artificial neural network, Comput. Fluids 195, 104319 (2019).
  57. J. Shen, Efficient spectral-Galerkin method II. Direct solvers of second-and fourth-order equations using Chebyshev polynomials, SIAM J. Sci. Comput. 16, 74 (1995).
  58. K. Julien and M. Watson, Efficient multi-dimensional solution of PDEs using Chebyshev spectral methods, J. Comput. Phys. 228, 1480 (2009).
  59. B. Sanderse, P. Stinis, R. Maulik, and S. E. Ahmed, Scientific machine learning for closure models in multiscale problems: A review, Found. Data Sci. 7, 298 (2025).
  60. P. Sagaut, Large Eddy Simulation for Incompressible Flows (Springer, Berlin, Heidelberg, 2006).
  61. S. Ghosal and P. Moin, The basic equations for the large eddy simulation of turbulent flows in complex geometry, J. Comput. Phys. 118, 24 (1995).
  62. G. R. Yalla, T. A. Oliver, S. W. Haering, B. Engquist, and R. D. Moser, Effects of resolution inhomogeneity in large-eddy simulation, Phys. Rev. Fluids 6, 074604 (2021).
  63. K. Jakhar, Y. Guan, R. Mojgani, A. Chattopadhyay, and P. Hassanzadeh, Learning closed-form equations for subgrid-scale closures from high-fidelity data: Promises and challenges, J. Adv. Model. Earth Syst. 16, e2023MS003874 (2024).
  64. S. Ouala, B. Chapron, F. Collard, L. Gaultier, and R. Fablet, Online calibration of deep learning sub-models for hybrid numerical modeling systems, Commun. Phys. 7, 402 (2024).
  65. H. Frezat, R. Fablet, G. Balarac, and J. Le Sommer, Gradient-free online learning of subgrid-scale dynamics with neural emulators, J. Adv. Model. Earth Syst. 18, e2023MS004140 (2026).
  66. J. Bradbury, R. Frostig, P. Hawkins, M. J. Johnson, C. Leary, D. Maclaurin, G. Necula, A. Paszke, J. VanderPlas, S. Wanderman-Milne, and Q. Zhang, JAX: Composable transformations of Python+NumPy programs, GitHub, 2018, http://github.com/google/jax.
  67. A. G. Baydin, B. A. Pearlmutter, A. A. Radul, and J. M. Siskind, Automatic differentiation in machine learning: A survey, J. Mach. Learn. Res. 18, 1 (2017).
  68. F. Sapienza, J. Bolibar, F. Schäfer, B. Groenke, A. Pal, V. Boussange, P. Heimbach, G. Hooker, F. Pérez, P.-O. Persson, et al., Differentiable programming for differential equations: A review, arXiv:2406.09699.
  69. I. Loshchilov and F. Hutter, Decoupled weight decay regularization, in Proceeding of 7th International Conference on Learning Representations (ICLR 2019) (Curran Associates, Inc., New Orleans, LA, USA, 2019).
  70. A. Hirose, Complex-Valued Neural Networks (Springer, Berlin, Heidelberg, 2006).
  71. C. Trabelsi, O. Bilaniuk, Y. Zhang, D. Serdyuk, S. Subramanian, J. F. Santos, S. Mehri, N. Rostamzadeh, Y. Bengio, and C. J. Pal, Deep complex networks, in Proceeding of 6th International Conference on Learning Representations (ICLR 2018) (Curran Associates, Inc., Vancouver, BC, Canada, 2018).
  72. Z. Liu, H. Mao, C.-Y. Wu, C. Feichtenhofer, T. Darrell, and S. Xie, A ConvNet for the 2020s, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (IEEE, New Orleans, LA, USA, 2022), pp. 11976–11986.
  73. M. Arjovsky, A. Shah, and Y. Bengio, Unitary evolution recurrent neural networks, in International Conference on Machine Learning (PMLR, New York, NY, USA, 2016), pp. 1120–1128.
  74. A. Hjørungnes and D. Gesbert, Complex-valued matrix differentiation: Techniques and key results, IEEE Trans. Signal Process. 55, 2740 (2007).
  75. O. Shishkina, R. J. Stevens, S. Grossmann, and D. Lohse, Boundary layer structure in turbulent thermal convection and its consequences for the required numerical resolution, New J. Phys. 12, 075022 (2010).
  76. O. Barrois, T. Gastine, and C. C. Finlay, Comparison of quasi-geostrophic, hybrid and 3-D models of planetary core convection, Geophys. J. Int. 231, 129 (2022).
  77. B. Davy, C. J. Davies, J. E. Mound, and S. M. Tobias, Effect of hyperdiffusion on rotating Rayleigh-Bénard convection, Phys. Rev. Fluids 11, 013502 (2026).
  78. C. Leith, Stochastic models of chaotic systems, Physica D 98, 481 (1996).
  79. B. Fox-Kemper and D. Menemenlis, Can large eddy simulation techniques improve mesoscale rich ocean models? Geophys. Monogr. Ser. 177, 319 (2008).
  80. S. D. Bachman, B. Fox-Kemper, and B. Pearson, A scale-aware subgrid model for quasi-geostrophic turbulence, J. Geophys. Res.: Oceans 122, 1529 (2017).
  81. I. Grooms, Backscatter in energetically-constrained Leith parameterizations, Ocean Modell. 186, 102265 (2023).
  82. T. Wilder and T. Kuhlbrodt, Examining the fidelity of Leith subgrid closures for parameterizing mesoscale eddies in idealized and global (NEMO) ocean models, J. Adv. Model. Earth Syst. 17, e2025MS004950 (2025).
  83. P. B. Rhines, Waves and turbulence on a beta-plane, J. Fluid Mech. 69, 417 (1975).
  84. M. Heimpel and J. Aurnou, Turbulent convection in rapidly rotating spherical shells: A model for equatorial and high latitude jets on Jupiter and Saturn, Icarus 187, 540 (2007).
  85. M. E. McIntyre, Potential vorticity, Encyclopedia of Atmospheric Sciences 2, 685 (2003).
  86. R. K. Scott and D. G. Dritschel, The structure of zonal jets in geostrophic turbulence, J. Fluid Mech. 711, 576 (2012).
  87. J. Verhoeven and S. Stellmach, The compressional beta effect: A source of zonal winds in planets? Icarus 237, 143 (2014).
  88. S. Sukoriansky, B. Galperin, and N. Dikovskaya, Universal spectrum of two-dimensional turbulence on a rotating sphere and some basic features of atmospheric circulation on giant planets, Phys. Rev. Lett. 89, 124501 (2002).
  89. B. Galperin, S. Sukoriansky, and N. Dikovskaya, Geophysical flows with anisotropic turbulence and dispersive waves: Flows with a β-effect, Ocean Dyn. 60, 427 (2010).
  90. G. Boffetta and R. E. Ecke, Two-dimensional turbulence, Annu. Rev. Fluid Mech. 44, 427 (2012).
  91. S. Sukoriansky, N. Dikovskaya, and B. Galperin, On the arrest of inverse energy cascade and the Rhines scale, J. Atmos. Sci. 64, 3312 (2007).
  92. B. Galperin, R. M. Young, S. Sukoriansky, N. Dikovskaya, P. L. Read, A. J. Lancaster, and D. Armstrong, Cassini observations reveal a regime of zonostrophic macroturbulence on Jupiter, Icarus 229, 295 (2014).
  93. S. Cabanes, T. Gastine, and A. Fournier, Zonostrophic turbulence in the subsurface oceans of the Jovian and Saturnian moons, Icarus 415, 116047 (2024).
  94. R. Wordsworth, P. Read, and Y. Yamazaki, Turbulence, waves, and jets in a differentially heated rotating annulus experiment, Phys. Fluids 20, 126602 (2008).
  95. T. MacRobert, XVI.—Fourier integrals, Proc. R. Soc. Edinburgh 51, 116 (1932).
  96. G. Cinelli, An extension of the finite Hankel transform and applications, Int. J. Eng. Sci. 3, 539 (1965).
  97. G. Boffetta, Energy and enstrophy fluxes in the double cascade of two-dimensional turbulence, J. Fluid Mech. 589, 253 (2007).
  98. D. Lemasquerier, B. Favier, and M. Le Bars, Zonal jets at the laboratory scale: Hysteresis and Rossby waves resonance, J. Fluid Mech. 910, A18 (2021).
  99. E. Simonnet, J. Rolland, and F. Bouchet, Multistability and rare spontaneous transitions in barotropic β-plane turbulence, J. Atmos. Sci. 78, 1889 (2021).
  100. I. J. Shokar, R. R. Kerswell, and P. H. Haynes, Stochastic latent transformer: Efficient modeling of stochastically forced zonal jets, J. Adv. Model. Earth Syst. 16, e2023MS004177 (2024).
  101. P. Marti, M. Calkins, and K. Julien, A computationally efficient spectral method for modeling core dynamics, Geochem. Geophys. Geosyst. 17, 3031 (2016).
  102. H. Frezat, GitHub, 2026, https://github.com/hrkz/online-sgs-qg-planetary.
  103. H. Frezat, hrkz/online-sgs-qg-planetary: [Computer software], Zenodo, 2026, doi: 10.5281/zenodo.22915271.
  104. J. P. Boyd, Chebyshev and Fourier Spectral Methods, 2nd ed. (Dover, Mineola, 2001).
  105. C. Canuto, M. Y. Hussaini, A. Quarteroni, and T. A. Zang, Spectral Methods: Fundamentals in Single Domains, Scientific Computation (Springer, Berlin, Heidelberg, 2006), Vol. 285.
  106. G. A. Glatzmaier, Numerical simulations of stellar convective dynamos. I. The model and method, J. Comput. Phys. 55, 461 (1984).

Outline

Information

Phys. Rev. Fluids 11, 104904– Published 14 October, 2026
  • Received 27 July 2026
  • Accepted 21 September 2026
Crossmark - Check for updates button

©2026 American Physical Society

Sign In to Your Journals Account

Filter

Filter

Article Lookup

Enter a citation