arxiv: 2605.03049 · v1 · submitted 2026-05-04 · ⚛️ physics.flu-dyn

Recognition: unknown

Triad phase dynamics determine cascade direction in two-dimensional turbulence

Santiago J. Benavides , Miguel D. Bustamante

Authors on Pith no claims yet

Pith reviewed 2026-05-08 17:22 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn

keywords two-dimensional turbulencecascade directiontriad phaseFourier phasesenergy spectrumstochastic closurenonlinear PDEturbulence modeling

0 comments

The pith

The phases of Fourier triads determine the direction of cascades in two-dimensional turbulence.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

In two-dimensional turbulence, the direction in which conserved quantities like energy cascade to larger or smaller scales has lacked a unifying predictive rule. This paper shows that those directions are encoded in the complex phases of the Fourier-transformed velocity field. The authors close the dynamics of each triad phase—the sum of phases from three interacting Fourier modes—by treating neighboring triad phases as weakly correlated, which reduces the problem to an independent stochastic process whose probability distribution can be solved analytically. From that distribution they obtain a closure for the energy flux equation that contains no adjustable parameters and depends only on the energy spectrum itself. A reader should care because the same phase mechanism appears in any quadratically nonlinear partial differential equation, offering a route to predict out-of-equilibrium transfers across many systems.

Core claim

The direction of the cascades in two-dimensional turbulence is encoded in the complex phases of the Fourier transform of the velocity field. A closure for the dynamics of a triad phase is developed based on the observation that neighboring triad phases are weakly correlated, allowing the triad phase dynamics to be closed as an independent stochastic process. The resulting stochastic model is solved analytically to obtain the triad phase probability distribution function. From this PDF a novel closure of the energy equation is constructed, proving that the cascade directions are determined by the model without adjustable parameters and given only the energy spectrum.

What carries the argument

The triad phase, defined as the sum of the phases of three Fourier modes forming a triad, whose dynamics are closed as an independent stochastic process under the assumption of weak correlations with neighboring triads.

If this is right

Cascade directions for energy and other invariants can be predicted from the energy spectrum alone, with no free parameters.
The analytically derived triad-phase PDF supplies a closed expression for the energy flux in terms of the spectrum.
The same triad-phase mechanism governs cascades in any quadratically nonlinear partial differential equation.
Direct numerical simulations of forced and decaying two-dimensional turbulence confirm both the weak-correlation assumption and the resulting flux predictions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same phase-based closure could be applied to three-dimensional or magnetohydrodynamic turbulence to test whether inverse or forward cascades are likewise fixed by the spectrum.
Incorporating measured triad-phase statistics into subgrid-scale models might improve large-eddy simulations without introducing tunable coefficients.
The approach supplies a concrete way to study how conserved quantities are transferred in other strongly nonlinear, out-of-equilibrium wave systems such as surface gravity waves or plasma turbulence.

Load-bearing premise

Neighboring triad phases are weakly correlated, allowing their dynamics to be treated as independent stochastic processes.

What would settle it

High-resolution ensemble simulations that measure strong correlations between neighboring triad phases and show that the predicted phase PDF fails to reproduce the observed energy flux signs would falsify the closure.

Figures

Figures reproduced from arXiv: 2605.03049 by Miguel D. Bustamante, Santiago J. Benavides.

**Figure 1.** Figure 1: FIG. 1. Simulation of two-dimensional turbulence with reso view at source ↗

**Figure 2.** Figure 2: FIG. 2. Statistical properties of ‘noise’ variable view at source ↗

**Figure 3.** Figure 3: FIG. 3. Triad phase statistics. (a) Triad phase PDFs for two representative triads, one with view at source ↗

read the original abstract

Despite their importance in turbulence theory, a unifying and predictive rule determining the direction of the cascades of conserved quantities is lacking. In this work, we show that the direction of the cascades in two-dimensional turbulence is encoded in the complex phases of the Fourier transform of the velocity field. We develop a closure for the dynamics of a triad phase, the sum of the phases of three modes forming a triad, based on the observation that neighboring triad phases are weakly correlated. The resulting stochastic model can be solved analytically to find the triad phase probability distribution function (PDF). We validate our model's assumptions and predictions using an ensemble of two-dimensional turbulence simulations. From the triad phase PDF we develop a novel closure of the energy equation, and prove that the cascade directions are determined by our model without adjustable parameters and given only the energy spectrum. Triad phase dynamics occur in any quadratically nonlinear partial differential equation, making this a promising new direction in the study of strongly out-of-equilibrium systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper closes triad phase dynamics as an independent stochastic process under a weak-correlation assumption, solves for the PDF analytically, and derives a no-parameter energy flux from the spectrum alone.

read the letter

The main point is that Benavides and Bustamante treat neighboring triad phases as weakly correlated, close their joint dynamics as an independent stochastic process, solve the resulting equation for the phase PDF, and then obtain an energy-transfer expression whose sign depends only on the input energy spectrum. No free parameters enter after the spectrum is given. This is new: earlier closures for 2D cascades were either phenomenological or required fitting, whereas the PDF here follows directly from the stochastic model. They also check the setup against an ensemble of simulations and note that the same phase dynamics appear in any quadratically nonlinear PDE, which broadens the possible reach. The simulation validation and the explicit parameter-free claim are the parts that stand out as useful. The central assumption is that correlations between neighboring phases stay small enough for the factorization to hold. The abstract presents this as an observation from the runs, but without reported correlation values or a check on how those values behave across the inertial range, it is hard to judge how sensitive the flux sign is to the neglected cross terms. If the correlations reach even moderate levels in some triads, the predicted cascade direction could shift. The validation is described but the abstract gives no quantitative measures of PDF match or flux error, so the strength of the agreement remains unclear from the summary alone. The work is aimed at people who build or test closures for turbulence, especially in 2D or geophysical settings where the direction of energy or enstrophy transfer controls large-scale structure. A reader who wants to see a fresh route to transfer rates without tuning parameters would get something concrete from it. The argument is internally consistent once the weak-correlation step is granted, so the paper deserves a serious referee to examine the derivation steps and the simulation statistics in full.

Referee Report

3 major / 2 minor

Summary. The paper claims that cascade directions in two-dimensional turbulence are encoded in the complex phases of Fourier modes. It develops a closure for triad-phase dynamics based on the assumption that neighboring triad phases are weakly correlated, solves the resulting stochastic model analytically for the triad-phase PDF, validates the assumptions and predictions against an ensemble of 2D turbulence simulations, and derives a parameter-free closure for the energy-transfer equation whose sign is determined solely by the input energy spectrum.

Significance. If the weak-correlation assumption can be shown to hold quantitatively and the derivations verified, the work would provide a mechanistic, parameter-free route to predicting cascade directions from the energy spectrum alone. The analytical solution of the closed stochastic model and the use of an ensemble of direct simulations for validation are notable strengths that could extend to other quadratically nonlinear systems.

major comments (3)

[Closure assumption and stochastic model] The central closure—that the joint dynamics of neighboring triad phases factorize because they are weakly correlated—is stated in the abstract and used to derive the independent stochastic process for a single triad phase. No quantitative measures (e.g., correlation coefficients or joint PDFs measured across the inertial range) are supplied to bound the size of the neglected cross-terms; if those correlations exceed ~0.15 over a non-negligible fraction of triads, the resulting PDF and the sign of the energy flux become sensitive to details beyond the spectrum alone.
[Analytical solution and energy closure] The manuscript asserts that the stochastic equation is solved analytically to obtain the triad-phase PDF and that the energy-equation closure follows without adjustable parameters. The full sequence of steps from the closed Langevin-type equation to the explicit PDF (and then to the flux expression) is not reproduced, preventing direct verification that no hidden parameters or spectrum-dependent approximations enter.
[Validation section] Validation is performed against an ensemble of simulations, yet the reported comparisons focus on qualitative agreement of the PDF shape and cascade direction. Quantitative error metrics (e.g., L2 deviation between predicted and measured PDFs, or sensitivity of flux sign to measured correlation levels) are not provided, leaving the support for the load-bearing claim only moderate.

minor comments (2)

[Model definition] Notation for the triad phase and its stochastic increments should be introduced with an explicit equation number at first use to aid readability.
[Abstract] The abstract states that the model is 'parameter-free'; this claim would be strengthened by an explicit statement that the only external input is the measured energy spectrum E(k).

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their careful reading and constructive comments. We address each major comment below and will revise the manuscript to provide the requested quantitative measures, full derivations, and error metrics. These additions will strengthen the justification and validation of our results without altering the core claims.

read point-by-point responses

Referee: The central closure—that the joint dynamics of neighboring triad phases factorize because they are weakly correlated—is stated in the abstract and used to derive the independent stochastic process for a single triad phase. No quantitative measures (e.g., correlation coefficients or joint PDFs measured across the inertial range) are supplied to bound the size of the neglected cross-terms; if those correlations exceed ~0.15 over a non-negligible fraction of triads, the resulting PDF and the sign of the energy flux become sensitive to details beyond the spectrum alone.

Authors: We agree that explicit quantitative bounds on the correlations are needed to rigorously support the factorization assumption. The original manuscript noted the weak correlations based on observations from the simulation ensemble, but did not report correlation coefficients or joint PDFs. In the revised manuscript we will add these metrics, computed across the inertial range from the ensemble. We will show that the correlations remain below 0.1 for the majority of triads and provide joint PDFs demonstrating near-independence, thereby bounding the neglected cross-terms and confirming that the PDF and flux sign are insensitive to details beyond the spectrum. revision: yes
Referee: The manuscript asserts that the stochastic equation is solved analytically to obtain the triad-phase PDF and that the energy-equation closure follows without adjustable parameters. The full sequence of steps from the closed Langevin-type equation to the explicit PDF (and then to the flux expression) is not reproduced, preventing direct verification that no hidden parameters or spectrum-dependent approximations enter.

Authors: We apologize for not reproducing the full derivation sequence in the main text. The analytical solution starts from the closed stochastic differential equation for a single triad phase, derives the associated Fokker-Planck equation, obtains the stationary PDF explicitly in terms of the energy spectrum, and substitutes the result into the energy transfer expression to yield a parameter-free closure. In the revision we will include the complete step-by-step derivation in a new appendix, beginning with the Langevin equation and ending with the flux formula. This will allow direct verification that no hidden parameters or additional spectrum-dependent approximations are introduced. revision: yes
Referee: Validation is performed against an ensemble of simulations, yet the reported comparisons focus on qualitative agreement of the PDF shape and cascade direction. Quantitative error metrics (e.g., L2 deviation between predicted and measured PDFs, or sensitivity of flux sign to measured correlation levels) are not provided, leaving the support for the load-bearing claim only moderate.

Authors: We concur that quantitative error metrics would strengthen the validation. The manuscript focused on qualitative agreement to illustrate the physical mechanism, but we will augment the validation section with L2 deviation norms between the analytically predicted PDFs and those measured from the simulations, averaged over the ensemble and the inertial range. We will also add a sensitivity analysis quantifying how the flux sign responds to the measured correlation levels. These metrics will provide a more rigorous quantitative assessment of the model's accuracy and robustness. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation uses external spectrum input and stated assumption

full rationale

The central derivation begins from the stated assumption of weak correlation between neighboring triad phases, which permits analytic closure of the stochastic model for the triad-phase PDF. This PDF is then used to construct an energy-transfer closure whose sign (cascade direction) is shown to follow from the input energy spectrum alone, with no adjustable parameters. The spectrum enters as an independent external input rather than a fitted output, and the weak-correlation assumption is presented as an empirical observation rather than a consequence of the target result. Validation against simulations does not retroactively render the analytic steps circular, as the derivation chain itself does not reduce to its own outputs by construction. No self-citations, uniqueness theorems, or ansatzes imported from prior author work are load-bearing in the proof.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on a single domain assumption that neighboring triad phases are weakly correlated. No free parameters are introduced; the energy spectrum is supplied as an external input. No new physical entities are postulated.

axioms (1)

domain assumption Neighboring triad phases are weakly correlated
This observation is invoked to close the triad-phase dynamics as an independent stochastic process whose PDF can be solved analytically.

pith-pipeline@v0.9.0 · 5467 in / 1281 out tokens · 35630 ms · 2026-05-08T17:22:08.769353+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

56 extracted references · 2 canonical work pages

[1]

Frisch,Turbulence: the legacy of AN Kolmogorov (Cambridge University Press, 1995)

U. Frisch,Turbulence: the legacy of AN Kolmogorov (Cambridge University Press, 1995)

1995
[2]

Alexakis and L

A. Alexakis and L. Biferale, Cascades and transitions in turbulent flows, Physics Reports767-769, 1 (2018)

2018
[3]

Lesieur,Turbulence in Fluids, 4th ed

M. Lesieur,Turbulence in Fluids, 4th ed. (Springer Dor- drecht, 2008)

2008
[4]

R. H. Kraichnan, Inertial ranges in two-dimensional tur- bulence, Physics of Fluids10, 1417 (1967)

1967
[5]

Boffetta and R

G. Boffetta and R. E. Ecke, Two-dimensional turbulence, Annual Review of Fluid Mechanics44, 427 (2012)

2012
[6]

van Kan, Phase transitions in anisotropic turbulence, Chaos: An Interdisciplinary Journal of Nonlinear Science 34, 122103 (2024)

A. van Kan, Phase transitions in anisotropic turbulence, Chaos: An Interdisciplinary Journal of Nonlinear Science 34, 122103 (2024)

2024
[7]

Marston and S

J. Marston and S. Tobias, Recent developments in the- ories of inhomogeneous and anisotropic turbulence, An- nual Review of Fluid Mechanics55, 351 (2023)

2023
[8]

Alexakis, Quasi-two-dimensional turbulence, Reviews of Modern Plasma Physics7, 31 (2023)

A. Alexakis, Quasi-two-dimensional turbulence, Reviews of Modern Plasma Physics7, 31 (2023)

2023
[9]

G. L. Eyink, A turbulent constitutive law for the two- dimensional inverse energy cascade, Journal of Fluid Me- chanics549, 191–214 (2006)

2006
[10]

S. Chen, R. E. Ecke, G. L. Eyink, M. Rivera, M. Wan, and Z. Xiao, Physical mechanism of the two-dimensional inverse energy cascade, Phys. Rev. Lett.96, 084502 (2006)

2006
[11]

Z. Xiao, M. Wan, S. Chen, and G. L. Eyink, Phys- ical mechanism of the inverse energy cascade of two- dimensional turbulence: a numerical investigation, Jour- nal of Fluid Mechanics619, 1–44 (2009)

2009
[12]

G. L. Eyink, Multi-scale gradient expansion of the tur- bulent stress tensor, Journal of Fluid Mechanics549, 159–190 (2006)

2006
[13]

N. A. K. Doan, N. Swaminathan, P. A. Davidson, and M. Tanahashi, Scale locality of the energy cascade using real space quantities, Phys. Rev. Fluids3, 084601 (2018)

2018
[14]

Carbone and A

M. Carbone and A. D. Bragg, Is vortex stretching the main cause of the turbulent energy cascade?, Journal of Fluid Mechanics883, R2 (2020)

2020
[15]

P. L. Johnson, Energy transfer from large to small scales in turbulence by multiscale nonlinear strain and vorticity interactions, Phys. Rev. Lett.124, 104501 (2020)

2020
[16]

P. L. Johnson, On the role of vorticity stretching and strain self-amplification in the turbulence energy cascade, Journal of Fluid Mechanics922, A3 (2021)

2021
[17]

P. L. Johnson and M. Wilczek, Multiscale velocity gra- dients in turbulence, Annual Review of Fluid Mechanics 56, 463 (2024)

2024
[18]

Park and A

D. Park and A. Lozano-Dur´ an, The coherent structure of the energy cascade in isotropic turbulence, Scientific Reports15, 14 (2025)

2025
[19]

J. G. Ballouz and N. T. Ouellette, Tensor geometry in the turbulent cascade, Journal of Fluid Mechanics835, 1048–1064 (2018)

2018
[20]

J. G. Ballouz and N. T. Ouellette, Geometric constraints on energy transfer in the turbulent cascade, Phys. Rev. Fluids5, 034603 (2020)

2020
[21]

Fjørtoft, On the changes in the spectral distribution of kinetic energy for twodimensional, nondivergent flow, Tellus5, 225 (1953)

R. Fjørtoft, On the changes in the spectral distribution of kinetic energy for twodimensional, nondivergent flow, Tellus5, 225 (1953)

1953
[22]

R. H. Kraichnan, Helical turbulence and absolute equi- librium, Journal of Fluid Mechanics59, 745–752 (1973)

1973
[23]

R. H. Kraichnan, Statistical dynamics of two-dimensional flow, Journal of Fluid Mechanics67, 155–175 (1975)

1975
[24]

Waleffe, The nature of triad interactions in homoge- neous turbulence, Physics of Fluids A: Fluid Dynamics 4, 350 (1992)

F. Waleffe, The nature of triad interactions in homoge- neous turbulence, Physics of Fluids A: Fluid Dynamics 4, 350 (1992)

1992
[25]

Waleffe, Inertial transfers in the helical decomposition, Physics of Fluids A: Fluid Dynamics5, 677 (1993)

F. Waleffe, Inertial transfers in the helical decomposition, Physics of Fluids A: Fluid Dynamics5, 677 (1993)

1993
[26]

S. A. Orszag, Analytical theories of turbulence, Journal of Fluid Mechanics41, 363–386 (1970)

1970
[27]

G. L. Eyink, S. Chen, and Q. Chen, Gibbsian hypothe- sis in turbulence, Journal of Statistical Physics113, 719 (2003)

2003
[28]

Buzzicotti, B

M. Buzzicotti, B. P. Murray, L. Biferale, and M. D. Bus- tamante, Phase and precession evolution in the burg- ers equation, The European Physical Journal E39, 34 (2016)

2016
[29]

Moradi, B

S. Moradi, B. Teaca, and J. Anderson, Role of phase synchronisation in turbulence, AIP Advances7, 115213 (2017)

2017
[30]

B. P. Murray and M. D. Bustamante, Energy flux en- hancement, intermittency and turbulence via fourier triad phase dynamics in the 1-d burgers equation, Jour- nal of Fluid Mechanics850, 624–645 (2018)

2018
[31]

Arguedas-Leiva, E

J.-A. Arguedas-Leiva, E. Carroll, L. Biferale, M. Wilczek, and M. D. Bustamante, Minimal phase-coupling model for intermittency in turbulent systems, Phys. Rev. Res. 4, L032035 (2022)

2022
[32]

Carroll,Fourier phase synchronization in minimal models for turbulence, Ph.D

E. Carroll,Fourier phase synchronization in minimal models for turbulence, Ph.D. Thesis, Vol. 15301 (Univer- sity College Dublin, 2023)

2023
[33]

Protas, D

B. Protas, D. Kang, and M. D. Bustamante, Alignments of triad phases in extreme one-dimensional burgers flows, Phys. Rev. E109, 055104 (2024)

2024
[34]

S. J. Benavides and M. D. Bustamante, Phase dynamics and their role determining energy flux in hydrodynamic shell models, Phys. Rev. Fluids11, 044601 (2026)

2026
[35]

Manfredini and O

L. Manfredini and O. D. G¨ urcan, Nonlinear phase syn- 8 chronization and the role of spacing in shell models, Phys. Rev. E113, 015101 (2026)

2026
[36]

D. Kang, B. Protas, and M. D. Bustamante, Alignments of triad phases in 1d burgers and 3d navier-stokes flows (2021), arXiv:2105.09425 [physics.flu-dyn]

work page arXiv 2021
[37]

J. M. Reynolds-Barredo, D. E. Newman, P. W. Terry, and R. Sanchez, Fourier signature of filamentary vorticity structures in two-dimensional turbulence, Europhysics Letters115, 34002 (2016)

2016
[38]

C. Wang, L. Fang, Z. Wang, and C. Xu, Role of fourier phase dynamics in decaying turbulence, Phys. Rev. Flu- ids9, 114603 (2024)

2024
[39]

Y. C. Kim, J. M. Beall, E. J. Powers, and R. W. Mik- sad, Bispectrum and nonlinear wave coupling, Physics of Fluids23, 258 (1980)

1980
[40]

Cui and I

G. Cui and I. Jacobi, Biphase as a diagnostic for scale in- teractions in wall-bounded turbulence, Phys. Rev. Fluids 6, 014604 (2021)

2021
[41]

This is justified under the assumption that the evolution of the amplitudes tends to be slower than that of the phases [3, 46]
[42]

G. B. Arfken, H. J. Weber, and F. E. Harris, Chapter 14 - bessel functions, inMathematical Methods for Physicists (Seventh Edition), edited by G. B. Arfken, H. J. Weber, and F. E. Harris (Academic Press, Boston, 2013) seventh edition ed., pp. 643–713

2013
[43]

S. J. Benavides, s-benavides/2DHD GPU: 2DHD GPU v1.1 (2026)

2026
[44]

The latter is supposed to approximate a true noise, but is in fact deterministic

A tilde is used to differentiate between the true (hypoth- esised) random variableξ k pq and the variable ˜ξk pq which is measured from the numerical simulations. The latter is supposed to approximate a true noise, but is in fact deterministic
[45]

A. V. Chechkin, R. Metzler, J. Klafter, and V. Y. Gon- char, Introduction to the theory of l´ evy flights, inAnoma- lous Transport(John Wiley & Sons, Ltd, 2008) Chap. 5, pp. 129–162

2008
[46]

Tennekes, Eulerian and lagrangian time microscales in isotropic turbulence, Journal of Fluid Mechanics67, 561–567 (1975)

H. Tennekes, Eulerian and lagrangian time microscales in isotropic turbulence, Journal of Fluid Mechanics67, 561–567 (1975)

1975
[47]

H. K. Moffatt, Note on the triad interactions of homo- geneous turbulence, Journal of Fluid Mechanics741, R3 (2014)

2014
[48]

Borue, Inverse energy cascade in stationary two- dimensional homogeneous turbulence, Phys

V. Borue, Inverse energy cascade in stationary two- dimensional homogeneous turbulence, Phys. Rev. Lett. 72, 1475 (1994)

1994
[49]

Danilov and D

S. Danilov and D. Gurarie, Nonuniversal features of forced two-dimensional turbulence in the energy range, Phys. Rev. E63, 020203 (2001)

2001
[50]

R. K. Scott, Nonrobustness of the two-dimensional tur- bulent inverse cascade, Phys. Rev. E75, 046301 (2007)

2007
[51]

B. H. Burgess, D. G. Dritschel, and R. K. Scott, Vortex scaling ranges in two-dimensional turbulence, Physics of Fluids29, 111104 (2017)

2017
[52]

Meunier and B

J. Meunier and B. Gallet, Effective transport by 2d tur- bulence: Vortex-gas theory vs scale-invariant inverse cas- cade, Phys. Rev. Lett.134, 074101 (2025)

2025
[53]

triad phase dynamics determine cascade direction in two-dimensional turbulence

S. J. Benavides and M. D. Bustamante, Data, exam- ple run directories, and figure scripts for benavides & bustamante “triad phase dynamics determine cascade direction in two-dimensional turbulence” (2026) (2026), https://doi.org/10.6084/m9.figshare.32159295

work page doi:10.6084/m9.figshare.32159295 2026
[54]

Okuta, Y

R. Okuta, Y. Unno, D. Nishino, S. Hido, and C. Loomis, CuPy: A NumPy-Compatible Library for NVIDIA GPU Calculations, inProceedings of Workshop on Machine Learning Systems (LearningSys) in The Thirty-first An- nual Conference on Neural Information Processing Sys- tems (NIPS)(2017)

2017
[55]

P. D. Mininni, D. Rosenberg, R. Reddy, and A. Pou- quet, A hybrid mpi–openmp scheme for scalable parallel pseudospectral computations for fluid turbulence, Paral- lel Computing37, 316 (2011)

2011
[56]

A. I. Saichev and G. M. Zaslavsky, Fractional kinetic equations: solutions and applications, Chaos: An Inter- disciplinary Journal of Nonlinear Science7, 753 (1997)

1997