On Phase Transition of ITG Turbulence in the Dimits shift
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: We're looking at the paper: "On Phase Transition of ITG Turbulence in the Dimits shift".
Mira: Here's the gist of the paper. The transition between turbulent and zonal flow dominated states in ion temperature gradient turbulence is investigated by varying the ion temperature gradient in nonlinear gyrokinetic simulations.
Kai: This study shows that the Dimits shift is associated with a phase transition between zonal and turbulence dominated regimes.
Mira: Regimes of Turbulence and Flow The investigation reveals three distinct regimes based on an order parameter defined as the fraction of the zonal to total free energy in the system.
Kai: These regimes are classified as follows: 1.
Paper summary: Kai: So we're looking at this paper, "On Phase Transition of ITG Turbulence in the Dimits shift." It investigates how you can switch between turbulent and zonal flow dominated states by changing the ion temperature gradient in nonlinear gyrokinetic simulations.
Mira: Exactly. The main idea here is that this shift between those two states—turbulence versus zonal flow—is associated with what they call a phase transition. They find that this happens when the ion temperature gradient crosses a certain threshold, and this transition can be seen as a change in the system's order parameter, which is defined as the fraction of zonal to total free energy in the system.
Lev: When you talk about that order parameter, you mean you're tracking how much of that zonal free energy is present versus all the total free energy available to the plasma? That sounds like a lot of bookkeeping for hardware simulation.
Kai: Right. And they break down their observations into three distinct regimes based on that fraction: a turbulent regime at high gradients, a zonal flow dominated regime with strongly reduced heat transport when you have low gradients, and then an intermediate regime where you see this switching between the two states intermittently.
Mira: That intermediate state is interesting because it suggests there's not just a sharp switch; it's more of an intermittent switching between the two conditions. This whole study is driven by how these different regimes behave when you vary that temperature gradient in nonlinear gyrokinetic simulations.
Lev: From my side, I’m thinking about how those simulation setups translate to actual hardware performance. If we were trying to run this on a real machine, the fidelity of those measurements would really depend on how slowly we change that temperature gradient over time.
Kai: That connects directly to their analysis of memory effects. They used chained simulations to look at this, and they found an asymmetric response which implies a hysteresis loop when you lower the temperature gradient and then bring it back up again.
Mira: Hysteresis suggests that the system has a memory of its past state, and it takes more free energy to break apart those zonal flows than it does to form them. They compare this mechanism to latent heat in crystal melting, which is a way of thinking about why that hysteresis occurs.
Lev: So when you're on real hardware, if you lower the gradient slowly enough, the simulation shows that you can get into a state where those zonal flows persist even at higher gradients later on. That’s something we have to account for in error correction overhead.
Kai: They also looked at the spectral signatures of this transition by examining the free energy wave number spectrum, and they found scale dependence between those regimes. In the turbulent regime, they see roughly two slopes joining around a value of k times rho i equals zero point five.
Paper summary: Mira: And in contrast, in the zonal flow dominated case, they see a strong concentration of that free energy at small k times rho i equals zero point zero five. That tells us where the energy is actually stored when things are flowing zonally versus when it's turbulent.
Lev: The difference between those two spectral features would be critical for diagnosing which physics is dominant in real transport measurements, and if we're measuring the right thing on our simulation setup.
Kai: And they also compared the turbulent spectra between independent and chained simulations, highlighting the importance of memory effects again. Chained simulations showed a more robust zonal dominated state where the zonal fraction reached values close to one while approaching the linear stability threshold.
Mira: That robustness in chained simulations is what makes them feel like they're getting closer to a true dynamical transition, even though they acknowledge that finite systems won't perfectly represent a real thermodynamic phase transition.
Lev: So if we want to use this as a diagnostic tool, the paper suggests using the large scale spectral slope parameter alpha to retrieve that hysteresis effect, which is useful because it gives us an indicator without needing to run those long chained simulations every time.
Kai: The conclusion they draw is that this Dimits shift thing should be re-interpreted as being analogous to a phase transition in a thermodynamic sense. They see two distinct dynamical states with the possibility of transitioning between them, even if finite systems can't perfectly capture that true thermodynamic transition.
Mira: What this really means for us is that we need to think about the system having these two clear dynamical states and the mechanism by which they switch between them when you vary those gradients. The minimization of zonal flow damping is also mentioned as an important aspect of how this behavior manifests in practice.
Lev: From an engineering standpoint, if we want confinement, minimizing that zonal flow damping sounds like a key design goal because it affects how effectively the turbulence is suppressed or sustained.
Kai: So to wrap up on the paper "On Phase Transition of ITG Turbulence in the Dimits shift," they are suggesting that this phenomenon isn't just some noise; it’s a genuine dynamical transition between turbulent and zonal flow states that exhibits hysteresis.
Mira: And what this implies for the physics is that we need to look at how those free energy fractions change as you move across the threshold of the Dimits shift region, because that’s where the system's fundamental behavior switches.
Lev: It sounds like a complex interplay between fluid dynamics and thermodynamics, which is exactly what we see when trying to model real confinement issues on our simulators.
Kai: Yeah, so if you're driving or cooking right now, remember that the physical systems around you are also exhibiting these kinds of transitions between different states depending on the conditions—just like this turbulence study does with temperature gradients.
Conclusion: Kai: So we're wrapping up on this paper about "On Phase Transition of ITG Turbulence in the Dimits shift." Basically, they’re looking at how you can switch between turbulent and zonal flow states just by changing the ion temperature gradient in nonlinear gyrokinetic simulations.
Mira: Right. The core idea is that this transition isn't just a smooth change; it acts like a phase transition, which means there are two distinct dynamical states the system settles into depending on where that gradient falls.
Lev: From where I sit, I’m thinking about how those two states translate to real hardware. If we were running this on actual machines, we'd need to see if those simulation results hold up when you introduce the noise and dissipation that real physics has.
Kai: Exactly. The authors are suggesting this switch is governed by an order parameter—that fraction of zonal free energy versus total free energy—and that’s what signals the transition point.
Mira: And they find that this transition involves hysteresis, which means the system remembers its past state, and it takes different amounts of energy to get into those two states depending on how you approach the change.
Lev: Hysteresis is a big thing for error correction because it suggests that if we try to drive the system back and forth across that threshold too fast, we might not recover the intended flow state correctly on real hardware.
Kai: So when they look at the overall picture, they’re arguing that this Dimits shift phenomenon is fundamentally a thermodynamic transition between two very different physical behaviors in the plasma.
Mira: It moves the discussion away from just turbulence being turbulence and zonal flows being zonal; it frames it as a phase change governed by energy landscapes.
Lev: If this analogy holds up, it means our models for how confinement works need to account for these distinct energy basins, not just one smooth crossover curve.
Kai: So the question is, can we actually build a model that captures that memory effect and those two separate states when we simulate things with finite resources?
Mira: That’s the next big hurdle. We have this solid theoretical framework now, but making it robust enough for real hardware testing is where the real work starts.
Laboratoire de Physique des Plasmas, CNRS, Ecole Polytechnique, Sorbonne University, Observatoire de Paris
physics.plasm-ph
Submitted: 2026-05-26
Updated: 2026-05-26
Comments: 17 pages, 15 figures
Journal ref: Plasma Phys. Control. Fusion 68 (2026) 095006
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 83/100
The gist: The transition between turbulent and zonal flow dominated states in ion temperature gradient turbulence is investigated by varying the ion temperature gradient in nonlinear gyrokinetic simulations.
Key concepts
- Regimes of Turbulence and Flow
- The investigation divides the system into three states based on how much free energy is in zonal flow versus total flow. These include a highly turbulent state at high gradients, a zonal flow dominated state with low heat transport at low gradients, and an intermediate switching regime where the system alternates between the two.
- Dimits Shift
- The Dimits shift is analyzed as being equivalent to a phase transition in thermodynamics. This means it represents a fundamental change in the system's behavior, moving between distinct dynamical states—specifically, turbulence and zonal flow dominance—rather than just a smooth change.
- Hysteresis
- Hysteresis is observed when the temperature gradient is lowered and then increased back up; the system exhibits an asymmetric response. This effect suggests that breaking apart existing zonal flows requires more free energy input than forming them, similar to latent heat during crystal melting.
Terminology
Summary
The transition between turbulent and zonal flow dominated states in ion temperature gradient turbulence is investigated by varying the ion temperature gradient in nonlinear gyrokinetic simulations. p. 0 This study shows that the Dimits shift is associated with a phase transition between zonal and turbulence dominated regimes. p. 0
Regimes of Turbulence and Flow
The investigation reveals three distinct regimes based on an order parameter defined as the fraction of the zonal to total free energy in the system. These regimes are classified as follows:
-
A turbulent regime at high gradients. p. 0
-
A zonal flow dominated regime with strongly reduced heat transport at low gradients. p. 0
-
An intermediate regime characterized by intermittent switching between these two states. p. 0
Mechanism of Transition and Hysteresis
The transition is analyzed using chained simulations to explore memory effects, finding an asymmetric response implying a hysteresis loop when the temperature gradient is lowered and then increased back. The mechanism of this hysteresis is analogous to latent heat in crystal melting, where it takes more free energy to break apart zonal flows than to form them. p. 0 The order parameter used is the zonal free energy fraction, denoted as Ξ. p. 9
Spectral Signatures of the Transition
The dynamics of the free energy wave-number spectrum reveal scale dependence between regimes. In the turbulent regime, the spectrum shows roughly two slopes joining around k x ρi = 0. p. 05. In contrast, the zonal flow dominated case exhibits a strong concentration of free energy at small k x, with a dominant contribution at k x ρi = 0. p. 805.
Role of Simulation Methodology
The comparison between independent and chained simulations highlights the importance of memory effects. Chained simulations show a more robust zonal dominated state with a zonal fraction reaching values close to unity while approaching the linear stability threshold. p. 0 The analysis of the large scale spectral slope parameter α suggests that hysteresis can be retrieved using this indicator. p. 0
Conclusion on Phase Transition
The study suggests a re-interpretation of the Dimits shift as being analogous to a phase transition
in the thermodynamical sense. The system exhibits two distinct dynamical states, with the possibility of a transition between them, even though finite systems will not properly represent a true thermodynamical phase transition. p. 6 The minimization of zonal flow damping is an important aspect of the behavior reported in this paper. p. 5
Acknowledgments
This work was granted access to the HPC resources of IDCS support unit from École Polytechnique, as well as was provided with computer and storage resources by GENCI at IDRIS thanks to the grant 2025-AD010514291R1 on the supercomputer Jean Zay’s the V100 partition. The authors gratefully acknowledge P. L. Guillon for useful discussions. p. 3 This work has been carried out within the framework of the EUROfusion Consortium, funded by the European Union via the Euratom Research and Training Programme (Grant Agreement No. 101052200—EUROfusion) and within the framework of the French Research Federation for Fusion Studies. The paper is organized as follows: in section II the local gyrokinetic equation that is solved by the GENE code is presented together with the typical parameters that are used. p. 1 In the next section III, the basic physics of the Dimits shift is reminded. p. 1 In section IV, particular emphasis is placed on chained simulations, in which the ion temperature gradient is gradually varied in both decreasing and increasing directions to explore the system’s behavior across the threshold. p. 1 From a physical perspective, this procedure can be interpreted as playing on the heating. p. 1 In section V, the influence of varying the ion temperature gradient on the turbulent spectra is compared between chained and independent series of simulations in section V. p. 1
References
[1] R. J. Groebner, W. Pfeiffer, F. P. Blau, K. H. Burrell, E. S. Fairbanks, R. P. Seraydarian, H. John, and R. E., Experimentally inferred ion thermal diffusivity profiles in the doublet iii tokamak: Comparison with neoclassical theory.
[2] B. A. Carreras, Progress in anomalous transport research in toroidal magnetic confinement devices, IEEE Transactions on Plasma Science 25, 1281. p. 0
[3] A. J. Wootton, B. A. Carreras, H. Matsumoto, K., McGuire, W. A., Peebles, C., Ritz, P., Terry, P., and S. J., Fluctuations and anomalous transport in tokamaks. p. 0
[4] B. Coppi, M. N. Rosenbluth, and R. Z., Instabilities due to temperature gradients in complex magnetic field configurations, Physics of Fluids 10, 582.
[5] A. M. Dimits, G., Bateman, M. A., Beer, B. I., Cohen, W., Dorland, G. W., Hammett, C., Kim, J. E., Kinsey, M., Kotschenreuther, A. H.-Kritzl,, L. L.-Lao,, J. Mandrekas,, W. M.-Nevins,, S. E.-Parker,, A. J.-Redd,, D. E.-Shumaker,, R., Sydora, and J., Weiland, Comparisons and physics basis of tokamak transport models and turbulence simulations. p. 0
[6] A. G. Peeters, F. Rath, R. Buchholz, Y. Camenen, J., Candy, F., Casson, S. R., Grosshauser, W. A., Hornsby, D., Strintzi,, and A., Weikl,, Gradient-driven fluxtube simulations of ion temperature gradient turbulence close to the non-linear threshold. p. 1
[7] A. Weikl, A. G. Peeters, F. Rath, S. R., Grosshauser, R., Buchholz, W. A., Hornsby, F., Seiferling,, and D., Strintzi,, Ion temperature gradient turbulence close to the finite heat flux threshold. p. 0
[8] J. W. Connor and H. R., A review of theories of the L-H transition, Plasma Physics and Controlled Fusion 42, R1. p. 0
[9] F. Wagner, G., Becker, K., Behringer, D., Campbell, A., Eberhagen,, W. Engelhardt,, G. Fussmann,, O. Gehre,, J Gernhardt,, G von Gierke,, G Haas, M Huang, F Karger, M Keilhacker, O Klüber,, M Kornherr, K Lackner, G Lisitano,, G. Lister, H. M., Mayer, D Meisel., E R Müller., H Murmann,, H Niedermeyer,, W Poschenrieder,, H Rapp,, H Röhr,, F Schneider,, G Siller, E Speth, A Stäbler., K. H Steuer,, G Venus, O Vollmer, and Z Yü, Regime of improved confinement and high beta in neutral-beam-heated divertor discharges of the ASDEX tokamak. p. 0
[10] K. H., Burrell, T. N., Carlstrom, E J., Doyle, D Finkenthal,, P Gohil, R J Groebner,, D L Hillis,, J Kim, H Matsumoto,, R A Moyer, T H Osborne,, C L Rettig,, W A Peebles, T L Rhodes,, H S John, R D Stambaugh,, M. R Wade, and J G Watkins, Physics of the L-mode to H-mode transition in tokamaks. p. 2
[11] M. A. Malkov and P H., Diamond, Weak hysteresis in a simplified model of the L–H transition, Physics of Plasmas 16, 012504. p. 0
[12] K. Itoh, S.-I. Itoh, K Ida, S Inagaki,, Y Kamada,, K Kamiya,, J Q Dong,, C Hidalgo, T Evans, W H Ko., H Park, T Tokuzawa,, S Kubo, T Kobayashi,, Y Kosuga,, M Sasaki, G. S. Yun., S D Song,, N Kasuya, Y Nagashima, C Moon,, M Yoshinuma, R Makino,, T Tsujimura,, H Tsuchiya, and U Stroth., Hysteresis and fast timescales in transport relations of toroidal plasmas. p. 0
[13] E. Gravier, M. Lesur, T Reveille, T Drouot, and J Médina,, Transport hysteresis and zonal flow stimulation in magnetized plasmas. p. 0
[14] F. Grander, F. F., Locker, and A Kendl,, Hysteresis in the gyrofluid resistive drift wave turbulence to zonal flow transition, Physics of Plasmas 31, 052301. p. 0
[15] P. L. Guillon and Ö. D., Phase transition from turbulence to zonal flows in the Hasegawa–Wakatani system, Physics of Plasmas 32, 012306. p. 0
[16] F. Grander, T Gröfler, F. F., Locker, M Rinner,, and A Kendl,, Merging of zonal flows in gyrofluid resistive drift-wave turbulence (2026), arXiv:2602. p. 023007.
[17] F. Jenko, W Dorland, M Kotschenreuther, and B N., Electron temperature gradient driven turbulence, Physics of Plasmas 7, 1904. p. 0
[18] A J. Brizard and T S. Hahm, Foundations of nonlinear gyrokinetic theory, Reviews of Modern Physics 79, 421. p. 1
[19] M A Beer, Gyrofluid Models of Turbulent Transport in Tokamaks, Ph. p. 0D thesis, Princeton University (1994).
[20] J Candy, R E Waltz, S E Parker,, and Y Chen, Relevance of the parallel nonlinearity in gyrokinetic simulations of tokamak plasmas, Physics of Plasmas 13, 074501. p. 0
[21] P Morel, A B Navarro, M Albrecht-Marc,, D Carati,, F Merz,, T Görler,, and F Jenko. p. 0, Gyrokinetic large eddy simulations, Physics of Plasmas 18, 072301. p. 0
[22] P Morel, A B Navarro, M Albrecht-Marc,, D Carati,, F Merz,, T Görler,, and F Jenko. p. 0, Dynamic procedure for filtered gyrokinetic simulations, Physics of Plasmas 19, 012311. p. 0
[23] F Romanelli, Ion temperature-gradient-driven modes and anomalous ion transport in tokamaks, Physics of Fluids B: Plasma Physics 1, 1018. p. 0
[24] P H. Diamond, S. p. 8-I. Itoh, K Itoh,, and T S. Hahm., Zonal flows in plasma—a review, Plasma Physics and Controlled Fusion 47, R35. p. 0
[25] Ö. D Gürcan and P H. p. 0 Diamond., Zonal flows and pattern formation, Journal of Physics A: Mathematical and Theoretical 48, 293001. p. 2
[26] H Biglari, P H. Diamond, and P W Terry,, Influence of sheared poloidal rotation on edge turbulence, Physics of Fluids B: Plasma Physics 2, 1. p. 3
[27] M A Malkov, P H. Diamond, and M N Rosenbluth,, On the nature of bursting in transport and turbulence in drift wave–zonal flow systems, Physics of Plasmas 8, 5073. p. 0
[28] L Schmitz, L Zeng, T L Rhodes,, J C Hillesheim,, E J Doyle,, R J Groebner,, W A Peebles,, K H Burrell,, and G Wang, Role of zonal flow predator-prey oscillations in triggering the transition to H-mode confinement, Physical Review Letters 108, 155002. p. 7
[29] P Morel, Ö. p. 0 D Gürcan, and V Berionni,, Characterization of predator–prey dynamics using the evolution of free energy in plasma turbulence, Plasma Physics and Controlled Fusion 56, 015002. p. 0
[30] S Kobayashi, Ö. D Gürcan, and P H. p. 0 Diamond,, Direct identification of predator-prey dynamics in gyrokinetic simulations, Physics of Plasmas 22, 090702. p. 3
[31] Z Lin, T S. Hahm, W W Lee,, W M Tang,, and P H. Diamond., Effects of collisional zonal flow damping on turbulent transport, Physical Review Letters 83, 3645. p. 3
[32] M N Rosenbluth and F L Hinton, Poloidal flow driven by ion-temperature-gradient turbulence in tokamaks, Physical Review Letters 80, 724. p. 0
[33] M J Pueschel, T Dannert, and F Jenko., On the role of numerical dissipation in gyrokinetic vlasov simulations of plasma microturbulence, Computer Physics Communications 181, 1428. p. 3
[34] A A Schekochihin, S C Cowley, W Dorland,, G. W Hammett,, G G Howes,, G G Plunk,, E Quataert,, and T Tatsuno., Gyrokinetic turbulence: a nonlinear route to dissipation through phase space, Plasma Physics and Controlled Fusion 50, 124024. p. 1
[35] A B Navarro, P Morel, M Albrecht-Marc,, D Carati,, F Merz,, T Görler,, and F Jenko. p. 0, Free energy balance in gyrokinetic turbulence, Physics of Plasmas 18, 092307. p. 3
[36] G Dif-Pradalier, G Hornung, P Ghendrih,, Y Sarazin,, F Clairet,, L Vermare,, P H. Diamond, J Abiteboul,, T Cartier-Michaud, C Ehrlacher,, D Estève. p. 8, X Garbet,, V Grandgirard., Ö D Gürcan, P Hennequin, Y Kosuga, G Latu, P Maget, P Morel. p. 0, C Norscini,, R Sabot,, and A Storelli,, Finding the elusive E×B staircase in magnetized plasmas. p. 0
[37] M Lippert, F Rath, and A G Peeters., Size convergence of the E×B staircase pattern in flux-tube simulations of ion temperature gradient-driven turbulence, Physics of Plasmas 30, 074501. p. 0
[38] P L Guillon, Ö D Gürcan, G Dif-Pradalier, Y Sarazin,, and N Fedorczak,, Flux-driven turbulent transport using penalisation in the hasegawa–wakatani system, Journal of Plasma Physics 91, E145. p. 0
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI can model plasma confinement transitions by using zonal free energy fraction as an order parameter to characterize system states, allowing it to distinguish between
Zonal regime
(Zeta greater than 0.70),Intermediate regime
(30% less than Zeta less than 70%), andTurbulent regime
(less than 30%). -
Improved AI can predict the hysteresis loop in the response of zonal free energy to temperature gradient changes by analyzing chained simulations, which show a
path dependence of the system and a possible hysteresis in the response of zonal free energy to temperature gradient.
-
Improved AI can characterize spectral slope scaling using power law fits for large scale k x spectra, identifying that
the large scale spectrum in the zonal regime is generally steeper than the turbulent regime
by comparing exponents alpha values. -
Improved AI can identify bistability in intermediate regimes by observing
intermittent switching between zonal and turbulent states
and its manifestation as avery large standard deviation
in the order parameter Zeta, suggestingtwo distinct 'thermodynamical phases'.
-
Improved AI can utilize memory effects in simulations by performing chained simulations to investigate how the
persistence of such structures influence the evolution of the system,
noting that zonal flow contribution is higher in chained simulations for lower temperature gradients due tothe persistence of zonal flows that are generated during previous steps.
Sources
Related papers
- Implementation and verification of a new free boundary module for Resistive Wall Modes simulations with the nonlinear 3D full MHD code SpeCyl
- System Size Dependence of Collisionless Reconnection Rate
- SPECTRE: A robust solver for 3D equilibria with arbitrary topology
- High-quality electron beam generation from laser wakefield accelerators for driving compact free electron lasers
- Deflation Techniques for Stellarator Equilibrium and Optimization
- Rapid Vacuum Response Calculations for Non-axisymmetric Plasma Geometries