Pith. sign in

REVIEW 3 major objections 7 minor 29 references

Integrable KP-I equation thermalizes despite integrability

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · glm-5.2

2026-07-08 15:52 UTC pith:LWOANZFK

load-bearing objection KP-I wave kinetic equation thermalizes to generalized Rayleigh-Jeans states; compact support theorem is the real new result the 3 major comments →

arxiv 2607.06119 v1 pith:LWOANZFK submitted 2026-07-07 nlin.CD nlin.SI

Wave Kinetics and Thermalization in Kadomtsev-Petviashvili-I System

Kiran Venkata Kolluru , Giorgio Krstulovic , Alan C Newell , Sergey Nazarenko This is my paper
classification nlin.CD nlin.SI
keywords wave turbulenceintegrable systemsthermalizationRayleigh-Jeans distributionKP-I equationwave kinetic equationinverse scattering transformBose-Einstein condensation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper argues that the Kadomtsev-Petviashvili-I (KP-I) equation, despite being integrable via the inverse scattering transform, admits a nontrivial wave-kinetic equation whose solutions evolve toward generalized Rayleigh-Jeans equilibrium states. The central mechanism is a class of resonant three-wave interactions that redistribute energy in Fourier space while preserving infinitely many conserved quantities. The authors prove that these interactions confine wave action to a compact region of Fourier space, avoiding the ultraviolet catastrophe that typically plagues equilibrium statistical descriptions of classical waves. Numerical simulations of the wave-kinetic equation confirm thermalization: entropy plateaus, the collision integral vanishes, and the spectrum converges to the predicted Rayleigh-Jeans form with relative errors as small as 1e-12.

Core claim

The discovery is that weak integrability, as opposed to strong integrability, permits genuine resonant energy transfer between Fourier modes. For KP-I, the resonant manifold has a degenerate structure yielding infinitely many conserved quantities of the form F(ξ)−F(η), and the equilibrium spectrum takes the form n = 1/(F(ξ)−F(η)). The geometric structure of allowed triadic interactions restricts spectral spreading to a finite domain, a property the authors call self-truncation. The thermalized state features strong low-wavenumber peaks resembling Bose-Einstein condensation, and the authors conjecture that when the weak-nonlinearity assumption breaks down at these peaks, coherent lump solitso

What carries the argument

Wave kinetic equation

Load-bearing premise

The wave-kinetic equation is derived under assumptions of random phases and weak nonlinearity, yet the thermalized Rayleigh-Jeans states develop strong low-wavenumber peaks where those assumptions self-consistently break down.

What would settle it

If the wave-kinetic equation for KP-I fails to thermalize to Rayleigh-Jeans states for generic initial conditions, or if the compact-support theorem is violated by resonant interactions escaping the predicted domain.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. This paper studies the wave-kinetic equation (WKE) derived from the integrable Kadomtsev-Petviashvili-I (KP-I) equation. The authors prove that compactness of Fourier-space supports is preserved under the WKE dynamics, owing to the structure of the resonant manifold (Theorems 1, 4, 6, 7). They then numerically simulate the WKE and present evidence that solutions thermalize to generalized Rayleigh-Jeans (RJ) equilibrium states, using three diagnostics: entropy plateau, decay of the collision integral N_infty, and direct RJ fitting with relative errors of order 1e-9 to 1e-12. The thermalized states feature strong low-wavenumber peaks formed via nonlocal spectral transfer, which the authors note is akin to Bose-Einstein condensation. The paper also discusses the connection to lump solutions of KP-I and the breakdown of the WKE near the peaks.

Significance. The paper makes a genuine contribution by demonstrating that the WKE derived from an integrable PDE (KP-I) admits nontrivial dynamics with thermalization, which is conceptually important for the interface between integrability and statistical mechanics. The compact support theorems are cleanly proved from the resonant manifold structure and provide a rigorous foundation for the self-truncation property, which avoids the ultraviolet catastrophe. The numerical WKE simulations conserve invariants to relative error ~1e-12 and employ three independent thermalization diagnostics, which is commendable rigor. The RJ fitting procedure using a single horizontal slice is an elegant and falsifiable diagnostic. The paper is honest about the limitations, explicitly acknowledging that the WKE breaks down near the low-k peaks of the thermalized state.

major comments (3)
  1. §V (Summary and Discussion): The title and abstract state that thermalization is observed 'in the Kadomtsev-Petviashvili-I System,' but what is actually demonstrated is thermalization of the WKE derived from KP-I. The WKE (Eqs. 5-6) is obtained under the standard wave turbulence closure (random phases, weak nonlinearity, phase mixing). For an integrable PDE, phase mixing is not guaranteed, and the paper relies on Zakharov's classification of KP-I as 'weakly integrable' [11] to justify the existence of a nontrivial WKE. Fig. 1 provides a single DNS snapshot of KP-I showing beam-like structures that qualitatively resemble WKE predictions, but no quantitative comparison of the wave-action spectrum n(k,t) between KP-I DNS and WKE simulations is presented at any point in the evolution. The authors themselves acknowledge (§V) that the WKE 'must inevitably break down near these peaks' that are,
  2. §IV, Fig. 4 and surrounding text: The thermalization claim rests on N_infty reaching ~1e-11 (Run A) or ~1e-9 (Run B) at t/tau_nl ~ O(10^5). However, the paper does not discuss whether this level represents true convergence to zero or a plateau determined by numerical discretization (512x512 grid, trapezoidal rule, second-order Runge-Kutta with dt=2e-4). Given that the simulation runs to t/tau_nl ~ 10^5, accumulated numerical error could set the floor for N_infty. A convergence study with grid refinement, or at least a discussion of why the observed N_infty levels are above numerical noise rather than evidence of a residual non-equilibrium state, would strengthen the central claim.
  3. §IV, Eq. (19) and Fig. 4: The diagnostic N_infty is defined as a supremum over all resonant triads, but the paper does not specify how this supremum is computed numerically (e.g., over all grid-point triads, or a subset). Since N_infty is a supremum rather than an average, it is sensitive to outliers and to the behavior at domain boundaries. Clarifying the computational procedure and reporting what fraction of triads have bracket values near the supremum versus near zero would help assess whether the small N_infty values genuinely indicate pointwise convergence to the RJ state.
minor comments (7)
  1. §II, Eq. (4): The definition of n_k includes a factor of (L/2pi)^2 / k_x. The division by k_x is nonstandard and not motivated; a brief explanation of why this normalization is natural for KP-I would help the reader.
  2. Fig. 1 caption: The caption states 'after ~O(10^3) linear timescales corresponding to the mode k=(2,0),' but it is unclear whether this means 10^3 periods of the mode k=(2,0) or some other measure. Clarify the time unit.
  3. §IV: The nonlinear timescale tau_nl is defined as the time at which the maximum of the initial spectrum decreases to three-quarters of its initial value. This is a post hoc definition that depends on the dynamics itself. Consider reporting tau_nl in physical units or relative to the linear timescale for reproducibility.
  4. Appendix A: The KP-I DNS uses a 1024x1024 grid with dt=2e-4, but no information is given about the total simulation time, the de-aliasing effectiveness, or whether energy is conserved in the DNS. Since Fig. 1 is the only PDE evidence presented, basic conservation diagnostics for the DNS would be appropriate.
  5. Fig. 5 caption: The caption states 'The right panel shows the spectrum in the (kx, ky) plane, while the left panel displays it in the transformed (xi, eta) coordinates,' but the figure layout appears to show (kx, ky) on the left and (xi, eta) on the right, which is the reverse of the caption description.
  6. §V, Eq. (21): The lump solution formula has x' = x - 12(a^2 + b^2)t, but standard references give the lump velocity as proportional to (a^2 + b^2) with a different coefficient. Verify the numerical prefactor.
  7. References: Several author names have encoding issues (e.g., 'Universitè' instead of 'Université', 'Alan.C.Newell' missing spaces). Proofread the author affiliations and reference list.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for a careful and constructive report. The referee raises three major comments concerning: (1) the distinction between thermalization of the WKE versus thermalization of the KP-I PDE itself, and the absence of quantitative DNS–WKE comparison; (2) whether the observed N_infty levels represent true convergence or numerical artifacts; and (3) the computational procedure for evaluating the N_infty supremum. We address each below. We agree with the substance of Comments 1 and 3 and will revise the manuscript accordingly. For Comment 2, we provide a substantive defense of the existing evidence while also committing to additional convergence checks.

read point-by-point responses

  1. Referee: §V: The title and abstract state thermalization is observed 'in the Kadomtsev-Petviashvili-I System,' but what is actually demonstrated is thermalization of the WKE derived from KP-I. No quantitative DNS–WKE comparison is presented. The WKE closure assumptions (random phases, phase mixing) are not guaranteed for integrable PDEs. The authors acknowledge the WKE breaks down near the low-k peaks.

    Authors: The referee is correct that the title and abstract overstate what is demonstrated. The thermalization result pertains to the WKE, not to the KP-I PDE directly. We will revise the title and abstract to make this distinction explicit. Specifically, the title will be amended to 'Wave Kinetics and Thermalization in the Wave-Kinetic Equation of the Kadomtsev-Petviashvili-I System' (or similar), and the abstract will clarify that thermalization is observed in the WKE derived from KP-I, not in direct simulations of the KP-I PDE itself. We will also add a sentence in the introduction stating that a quantitative DNS–WKE comparison is an important open problem that lies beyond the scope of the present work. Regarding the WKE closure assumptions: we rely on Zakharov's classification of KP-I as 'weakly integrable' [11], which is the established theoretical basis for the existence of a nontrivial WKE for KP-I. The question of whether phase mixing actually occurs in KP-I DNS is indeed open, and we will state this more prominently. However, we note that Fig. 1 is presented as motivational, not as validation; the qualitative resemblance of beam-like structures is suggestive but we agree it does not constitute quantitative verification. We will adjust the framing of Fig. 1 accordingly. revision: yes

  2. Referee: §IV, Fig. 4: The N_infty levels (~1e-11 for Run A, ~1e-9 for Run B) may reflect numerical discretization floor rather than true convergence. No convergence study with grid refinement is presented. Given the long integration time (t/tau_nl ~ 1e5), accumulated numerical error could set the floor.

    Authors: We appreciate this concern. Several pieces of evidence support the interpretation that the observed N_infty levels reflect genuine convergence rather than a numerical artifact. First, the invariants are conserved to relative error ~1e-12 throughout the entire simulation (including at t/tau_nl ~ 1e5), which indicates that accumulated integration error remains well below the N_infty levels. If numerical error were setting the floor for N_infty, we would expect comparable degradation in invariant conservation. Second, the three diagnostics are independent: the entropy plateau, the N_infty decay, and the RJ fitting with relative errors of 1e-9 to 1e-12 all converge to consistent conclusions. The RJ fitting error (1e-9 to 1e-12) is particularly telling, as it is computed from a single horizontal slice and then validated against the full 2D spectrum — if the state were a residual non-equilibrium state rather than a genuine RJ equilibrium, the slice-based reconstruction would not match the full spectrum to this accuracy. Third, the different N_infty levels between Runs A and B (1e-11 vs 1e-9) are consistent with the different RJ fitting accuracies for the two runs (1e-12 vs 1e-11), suggesting that the residual reflects the quality of convergence rather than a uniform discretization floor. That said, we agree that a grid refinement study would strengthen the claim. We will add a convergence study with at least one higher-resolution run (1024x1024) and report the resulting N_infty levels. We will also add a discussion of why the observed levels are above numerical noise, referencing the invariant conservation as a cross-check. revision: partial

  3. Referee: §IV, Eq. (19): The N_infty supremum computation is not specified. How is the supremum computed numerically? What fraction of triads are near the supremum versus near zero? Sensitivity to outliers and boundary behavior is unclear.

    Authors: The referee is correct that the computational procedure for N_infty is not described in sufficient detail. We will add a paragraph in §IV (or Appendix A) specifying the procedure. In brief: N_infty is computed by evaluating the bracket |n_chi1 n_chi2 - n_chi n_chi1 - n_chi n_chi2| over all grid-point triads within the computational domain at each time step, and taking the maximum. We will also report the fraction of triads with bracket values near the supremum versus near zero, as the referee suggests. We note that the supremum is indeed sensitive to outliers, which is why we complement it with the entropy diagnostic (an integral quantity) and the RJ fitting procedure. The fact that all three diagnostics agree is the basis for the thermalization claim, not N_infty alone. We will make this complementarity explicit in the revised text and include the requested statistics on the distribution of bracket values across triads. revision: yes

Circularity Check

0 steps flagged

No significant circularity found; derivation is self-contained with standard external citations

full rationale

The paper's derivation chain is substantially self-contained. (1) The WKE (Eqs. 5-6) is derived from KP-I via standard wave turbulence theory, citing Zakharov [11] and standard WTT references [8-10] — none of which are self-citations in the load-bearing sense (Zakharov is not an author; Nazarenko's textbook [9] is a standard reference for the H-theorem, which is a textbook result, not a paper-specific claim). (2) The RJ solutions (Eq. 16) are stationary solutions of the WKE by construction, but the paper does not claim to 'predict' them — it claims to observe thermalization toward them numerically, supported by three independent diagnostics (entropy plateau, N_∞ → 0, and RJ fitting). (3) The compact support theorem (Theorem 1/4/6) is proved self-containedly in Appendix B from the resonant manifold structure (Eqs. 12-13), with no fitting or self-citation. (4) The RJ fitting procedure (Appendix A) reconstructs F(·) from a single horizontal slice of the final simulation data using the symmetry relation (Eq. 20), then tests whether n_RJ = 1/(F(ξ)-F(η)) matches the full 2D simulation data. This is a genuine cross-validation: the single-slice reconstruction does not determine n(ξ,η) everywhere by construction — the WKE dynamics could produce any spectrum, and the ~10^{-9} to 10^{-12} relative error confirms the RJ functional form. The only minor self-citation is Nazarenko's textbook [9] for the H-theorem, but this is a standard result independently verified by the numerical diagnostics, so it is not load-bearing. No step in the derivation reduces to its inputs by construction.

Axiom & Free-Parameter Ledger

3 free parameters · 5 axioms · 0 invented entities

The free parameters are all initial condition choices for numerical simulations, not fitted physical constants. The axioms are standard wave turbulence assumptions. No new physical entities are postulated. The WKE itself is derived from KP-I without introducing new physics.

free parameters (3)
  • Initial spectral support D0 = Square [1.5,4.5]x[-4.5,-1.5] (Run A); Disk radius 1 centered at (3,-3) (Run B)
    Chosen by the authors to define initial conditions; not fitted to physical data but selected for numerical demonstration.
  • Gaussian widths Delta_xi, Delta_eta = 0.8 (Run A), 1.0 (Run B)
    Set the initial spectral width; chosen ad hoc for localization.
  • Polynomial degree N for F(x) fit = 10
    Used for the polynomial approximation of the RJ functional form in Run A; chosen to achieve ~1e-2 accuracy.
axioms (5)
  • domain assumption Random phase approximation: phases of Fourier amplitudes are effectively random, enabling closure of the cumulant hierarchy
    Invoked in Section II to derive the WKE; standard in wave turbulence theory but an uncontrolled approximation for integrable systems.
  • domain assumption Weak nonlinearity limit: the WKE is valid in the limit of small wave amplitudes
    Section II; the authors note this breaks down near low-k peaks in the thermalized state (Section V).
  • domain assumption Large-box limit L->infinity taken before weak-nonlinearity limit
    Section II; standard WTT ordering assumption.
  • standard math H-theorem for the WKE
    Section II; cited from [8, 9], implies entropy is maximized by RJ distribution.
  • standard math Resonant manifold structure for KP-I has only two nontrivial solutions (Eq. 12-13)
    Section II; derived from the parametrization in [21, 22].

pith-pipeline@v1.1.0-glm · 22569 in / 2306 out tokens · 424013 ms · 2026-07-08T15:52:17.578889+00:00 · methodology

0 comments
read the original abstract

We study properties of solutions, both evolving and equilibrium of the wave-kinetic equation describing ensembles of weak random waves governed by the Kadomstev-Petviashivli-I equations. The latter equation is integrable by the inverse scattering method, and yet it allows resonant wave interactions leading to redistribution of energy in the Fourier space. Such resonant interactions preserve an infinite number of invariants and we find that they preserve compactness of Fourier space supports. Numerically, we observe that the system can thermalize to one of the equilibrium states of Rayleigh-Jeans type, despite the common empirical belief that thermalization is impossible for integrable systems. The thermalized states are formed via non-local spectral transfers leading to creation of strong low-wavenumber peaks of the wave spectrum -- a process akin to Bose-Einstein condensation.

Figures

Figures reproduced from arXiv: 2607.06119 by Alan C Newell, Giorgio Krstulovic, Kiran Venkata Kolluru, Sergey Nazarenko.

Figure 1
Figure 1. Figure 1: FIG. 1: Logarithmic color plot of the anisotropic Fourier amplitude spectrum [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: (a) Schematic of a compact domain [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Spectrum evolution in Run B. Left column: 2D color plot of [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Temporal evolution of bracket [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: Color log-scale plot of the wave-action spectrum corresponding to the superposition of two independent [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Schematic of a rectangular domain [PITH_FULL_IMAGE:figures/full_fig_p015_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7: Schematic of a compact domain [PITH_FULL_IMAGE:figures/full_fig_p019_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8: A schematic illustrating how the wave action spreads to the final support when a cut-point is involved. [PITH_FULL_IMAGE:figures/full_fig_p020_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9: A schematic illustrating how the wave action spreads to the final support when the sets are disjointed [PITH_FULL_IMAGE:figures/full_fig_p021_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10: A schematic illustrating how the wave action spreads to the final support when the initial disconnected [PITH_FULL_IMAGE:figures/full_fig_p022_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11: Temporal evolution of [PITH_FULL_IMAGE:figures/full_fig_p023_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12: Columns (I–III) show, respectively, the wave-action spectrum obtained from simulations at long times [PITH_FULL_IMAGE:figures/full_fig_p024_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13: For Run A Columns (I–III) show, respectively, the horizontal slice [PITH_FULL_IMAGE:figures/full_fig_p024_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14: Spectrum evolution in Run C. Left column: 2D color plot of [PITH_FULL_IMAGE:figures/full_fig_p025_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIG. 15: Evolution of bracket [PITH_FULL_IMAGE:figures/full_fig_p025_15.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages · 1 internal anchor

  1. [1]

    Studies of the nonlinear problems

    Enrico Fermi, P Pasta, Stanislaw Ulam, and Mary Tsingou. Studies of the nonlinear problems. Technical report, Los Alamos Scientific Lab., N. Mex., 1955

    work page 1955

  2. [2]

    SIAM, 1985

    Alan C Newell.Solitons in mathematics and physics. SIAM, 1985

    work page 1985

  3. [3]

    The fermi–pasta–ulam problem: fifty years of progress.Chaos: An Interdisciplinary Journal of Nonlinear Science, 15(1), 2005

    GP Berman and FM Izrailev. The fermi–pasta–ulam problem: fifty years of progress.Chaos: An Interdisciplinary Journal of Nonlinear Science, 15(1), 2005

    work page 2005

  4. [4]

    Springer Science & Business Media, 2007

    Giovanni Gallavotti.The Fermi-Pasta-Ulam problem: a status report, volume 728. Springer Science & Business Media, 2007

    work page 2007

  5. [5]

    Fermi, pasta, ulam, and a mysterious lady.Physics today, 61(1):55–57, 2008

    Thierry Dauxois. Fermi, pasta, ulam, and a mysterious lady.Physics today, 61(1):55–57, 2008

    work page 2008

  6. [6]

    solitons

    Norman J Zabusky and Martin D Kruskal. Interaction of" solitons" in a collisionless plasma and the recurrence of initial states.Physical review letters, 15(6):240, 1965

    work page 1965

  7. [7]

    Method for solving the korteweg-devries equation.Physical review letters, 19(19):1095, 1967

    Clifford S Gardner, John M Greene, Martin D Kruskal, and Robert M Miura. Method for solving the korteweg-devries equation.Physical review letters, 19(19):1095, 1967

    work page 1967

  8. [8]

    Springer Science & Business Media, 2012

    Vladimir E Zakharov, Victor S L’vov, and Gregory Falkovich.Kolmogorov spectra of turbulence I: Wave turbulence. Springer Science & Business Media, 2012

    work page 2012

  9. [9]

    Springer Science & Business Media, 2011

    Sergey Nazarenko.Wave turbulence, volume 825. Springer Science & Business Media, 2011

    work page 2011

  10. [10]

    Wave turbulence.Annual review of fluid mechanics, 43(1):59–78, 2011

    Alan C Newell and Benno Rumpf. Wave turbulence.Annual review of fluid mechanics, 43(1):59–78, 2011. 10 FIG. 5: Color log-scale plot of the wave-action spectrum corresponding to the superposition of two independent lump solutions of the KP-I equation. The right panel shows the spectrum in the(kx, ky)plane, while the left panel displays it in the transforme...

    work page 2011

  11. [11]

    Turbulence in integrable systems.Studies in Applied Mathematics, 122(3):219–234, 2009

    Vladimir E Zakharov. Turbulence in integrable systems.Studies in Applied Mathematics, 122(3):219–234, 2009

    work page 2009

  12. [12]

    On the stability of solitary waves in weakly dispersing media

    Boris Borisovich Kadomtsev and Vladimir Iosifovich Petviashvili. On the stability of solitary waves in weakly dispersing media. InDoklady Akademii Nauk, volume 192, pages 753–756. Russian Academy of Sciences, 1970

    work page 1970

  13. [13]

    Cambridge University Press, 2011

    Mark J Ablowitz.Nonlinear dispersive waves: asymptotic analysis and solitons, volume 47. Cambridge University Press, 2011

    work page 2011

  14. [14]

    Ontheevolutionofpacketsofwaterwaves.Journal of Fluid Mechanics, 92(4):691–715, 1979

    MarkJ.AblowitzandHarveySegur. Ontheevolutionofpacketsofwaterwaves.Journal of Fluid Mechanics, 92(4):691–715, 1979

    work page 1979

  15. [15]

    E. A. Kuznetsov and S. K. Turitsyn. Soliton instability and collapse in media with defocusing nonlinearity.Zhurnal Eksperimentalnoi i Teoreticheskoi Fiziki, 94:119–129, August 1988

    work page 1988

  16. [16]

    Route to thermalization in theα-fermi–pasta–ulam system.Proceedings of the National Academy of Sciences, 112(14):4208–4213, 2015

    Miguel Onorato, Lara Vozella, Davide Proment, and Yuri V Lvov. Route to thermalization in theα-fermi–pasta–ulam system.Proceedings of the National Academy of Sciences, 112(14):4208–4213, 2015

    work page 2015

  17. [17]

    Double scaling in the relaxation time in theβ-fermi-pasta-ulam-tsingou model.Physical review letters, 120(14):144301, 2018

    Yuri V Lvov and Miguel Onorato. Double scaling in the relaxation time in theβ-fermi-pasta-ulam-tsingou model.Physical review letters, 120(14):144301, 2018

    work page 2018

  18. [18]

    One-dimensional optical wave turbulence: experiment and theory.Physics Reports, 514(4):121–175, 2012

    Jason Laurie, Umberto Bortolozzo, Sergey Nazarenko, and Stefania Residori. One-dimensional optical wave turbulence: experiment and theory.Physics Reports, 514(4):121–175, 2012

    work page 2012

  19. [19]

    Optical wave turbulence and the condensation of light.Journal of the Optical Society of America B, 26(12):2280–2284, 2009

    Umberto Bortolozzo, Jason Laurie, Sergey Nazarenko, and Stefania Residori. Optical wave turbulence and the condensation of light.Journal of the Optical Society of America B, 26(12):2280–2284, 2009

    work page 2009

  20. [20]

    Wave turbulence in integrable systems: nonlinear propagation of incoherent optical waves in single-mode fibers.Optics express, 19(18):17852–17863, 2011

    Pierre Suret, Antonio Picozzi, and Stéphane Randoux. Wave turbulence in integrable systems: nonlinear propagation of incoherent optical waves in single-mode fibers.Optics express, 19(18):17852–17863, 2011

    work page 2011

  21. [21]

    Degenerative dispersion laws, motion invariants and kinetic equations.Physica D: Nonlinear Phenomena, 1(2):192–202, 1980

    VE Zakharov and EI Schulman. Degenerative dispersion laws, motion invariants and kinetic equations.Physica D: Nonlinear Phenomena, 1(2):192–202, 1980

    work page 1980

  22. [22]

    On additional motion invariants of classical hamiltonian wave systems.Physica D: Nonlinear Phenomena, 29(3):283–320, 1988

    VE Zakharov and EI Schulman. On additional motion invariants of classical hamiltonian wave systems.Physica D: Nonlinear Phenomena, 29(3):283–320, 1988

    work page 1988

  23. [23]

    Giorgio Krstulovic and Vincent Labarre. Wavkins. jl: an efficient and modular julia software for solving wave kinetic equations.arXiv preprint arXiv:2504.00252, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  24. [24]

    On the stability of nonlinear waves in integrable models.Physica D: Nonlinear Phenomena, 10(3):379–386, 1984

    Evegenii A Kuznetsov, MD Spector, and GE Fal’Kovich. On the stability of nonlinear waves in integrable models.Physica D: Nonlinear Phenomena, 10(3):379–386, 1984

    work page 1984

  25. [25]

    Cambridge university press, 2000

    Eryk Infeld and George Rowlands.Nonlinear waves, solitons and chaos. Cambridge university press, 2000

    work page 2000

  26. [26]

    Ablowitz, Gino Biondini, and Qiao Wang

    Mark J. Ablowitz, Gino Biondini, and Qiao Wang. Whitham modulation theory for the kadomtsev– petviashvili equation. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 473(2204):20160695, 08 2017. 11

    work page 2017

  27. [27]

    Condensation of Classical Nonlinear Waves.Physical Review Letters, 95(26):263901, December 2005

    Colm Connaughton, Christophe Josserand, Antonio Picozzi, Yves Pomeau, and Sergio Rica. Condensation of Classical Nonlinear Waves.Physical Review Letters, 95(26):263901, December 2005

    work page 2005

  28. [28]

    Pelinovsky, Yuri A

    Dmitry E. Pelinovsky, Yuri A. Stepanyants, and Yuri S. Kivshar. Self-focusing of plane dark solitons in nonlinear defocusing media.Phys. Rev. E, 51:5016–5026, May 1995

    work page 1995

  29. [29]

    Satsuma and M

    J. Satsuma and M. J. Ablowitz. Two-dimensional lumps in nonlinear dispersive systems.Journal of Mathematical Physics, 20(7):1496–1503, 07 1979. Appendix A: Numerical methods Direct numerical simulation of KP-1 equation We solve Eq. (1) withσ= 1using a pseudo-spectral method with periodic boundary conditions on a1024×1024grid ofcollocationpoints. Timeinteg...

    work page 1979