For the steady direct (enstrophy) cascade of two-dimensional Navier-Stokes turbulence, Falkovich and Lebedev (2011; detailed derivation with Stepanov in arXiv:1011.3647) obtained the first analytic probability-density-function (PDF) tail derived from the Navier-Stokes equation itself: the vorticity coarse-grained over a scale RR has a universal self-similar exponential tail ln⁡P≃−∣ϖ∣/[Hln⁡(L/R)]1/3\ln P \simeq -|\varpi|/[H\ln(L/R)]^{1/3}, determined solely by the enstrophy flux HH, up to an order-unity prefactor and a slow ln⁡ln⁡(L/R)\ln\ln(L/R) factor that the final answer drops - implying no power-law anomalous exponents. The authors state the result is untested quantitatively and the derivation only partially controlled. Numerical works confirm exponential tails only at order-of-magnitude level (Boffetta-Musacchio 2010; Kuznetsov-Sereshchenko 2017), while frictional runs report non-collapsing vorticity PDFs (Tan-Huang-Meng 2014); no study tests the order-unity prefactor, the double-logarithmic correction, or pumping universality; the assumed initial condition at the remote time t∗t_* and the subleading corrections remain untreated; the suggested extensions to the inverse 2D and 3D cascades are only begun (Kleineberg-Friedrich 2013; Schorlepp et al). The problem - establishing the tail law quantitatively by test, and placing its derivation on controlled footing - remains open.

Background

Two-dimensional (2D) turbulence is governed by the Navier-Stokes equation for the vorticity ω=∂xvy−∂yvx\omega = \partial_x v_y - \partial_y v_x, the curl of the velocity field. Because enstrophy (the mean squared vorticity) is conserved by the inviscid dynamics alongside energy, a flow stirred by a random force correlated on a length LL splits into two cascades: energy is transported upscale (inverse energy cascade) while enstrophy is transported downscale (the direct, or enstrophy, cascade), which occupies the scale interval between the forcing scale and the viscous scale η\eta. A central question of turbulence theory is universality: how much of the small-scale statistics is fixed by the flux alone, independent of the forcing details. In the inverse cascade the statistics is empirically scale invariant, while in known direct cascades self-similarity is generally broken (intermittency), and for Navier-Stokes turbulence it is not even known whether the exponents of the velocity moments are universal.

For the 2D direct cascade, Falkovich and Lebedev (published Rapid Communication, Phys. Rev. E 83, 045301(R) (2011); detailed derivation with M. Stepanov in preprint arXiv:1011.3647, cited by the published version as its ref [18]) derived the first PDF tail in Navier-Stokes turbulence obtained directly from the equation of motion. Writing ϖ\varpi for the vorticity coarse-grained (spatially averaged) over a disk of radius RR inside the direct-cascade interval, and HH for the mean enstrophy flux, they predict for strong fluctuations ϖ≫ϖrms=[Hln⁡(L/R)]1/3\varpi \gg \varpi_{\rm rms} = [H\ln(L/R)]^{1/3} the exponential tail

ln⁡P(ϖ)≃−∣ϖ∣[Hln⁡(L/R)]1/3,\ln P(\varpi) \simeq -\frac{|\varpi|}{[H\ln(L/R)]^{1/3}},

up to an order-unity prefactor and a slowly varying double-logarithmic factor ln⁡ln⁡(L/R)\ln\ln(L/R) that appears in their intermediate result (Eqs. (66)-(68) of the preprint) but is dropped in the final answer. The PDF is thus self-similar, P(ϖ)=ϖ−1f(ϖ3/[Hln⁡(L/R)])P(\varpi) = \varpi^{-1} f(\varpi^3/[H\ln(L/R)]): moments scale as [Hln⁡(L/R)]n/3[H\ln(L/R)]^{n/3} with no power-law anomalous exponents, making this the theory-side benchmark against which any claimed intermittency of the enstrophy cascade must be judged. Technically, the derivation works in the Martin-Siggia-Rose (MSR) path-integral representation of the stochastically forced vorticity equation, where correlation functions are functional integrals with a known effective action. A rotationally symmetric saddle point of this action (a 'naive instanton', the most probable field configuration realizing a prescribed rare fluctuation) is useless here because axial symmetry eliminates the vortex-stretching nonlinearity; the authors therefore first integrate over the angular harmonics of the vorticity in a Gaussian approximation (only the second harmonic contributes at leading logarithmic order in ln⁡(L/R)\ln(L/R)), obtaining a renormalized effective action for the rotationally symmetric part, whose saddle point is a 'slow' instanton whose vorticity grows linearly in time over the lifetime ∣t∗∣|t_*|.

The authors state the limitations explicitly. First, the supporting simulations (Pasquero-Falkovich 2002; Boffetta-Musacchio 2010) show only approximately exponential tails, and 'data with different pumping statistics and better resolution are needed for a quantitative comparison' - the prefactor and the ln⁡ln⁡(L/R)\ln\ln(L/R) correction are untested. Second, the initial condition at the remote time t∗t_*, where the saddle-point solution departs from the rms level - the second angular harmonic of the vorticity correlation - is adopted in a form consistent with the flux relation but, in the authors' words, 'cannot be derived from the equation of motion' within their approximation. Third, subleading terms are controlled only at leading-logarithmic accuracy: the neglect of the cubic interactions of the nonzero angular harmonics and of the higher cumulants in the harmonic integration is justified a posteriori at logarithmic accuracy, with the applicability condition ∣ϖ3∣≫Hln⁡(L/R)|\varpi^3| \gg H\ln(L/R) (Eq. (70) of the preprint), and the precise parameter domain where the saddle-point approximation is valid remains undelimited. Fourth, they state that the slow-variable method is unlikely to apply to cascades whose strong fluctuations are fast (such as the 3D energy cascade) and propose as further tasks its application to other temporally decelerating cascades, such as the 2D inverse cascade and wave-turbulence cascades.

Problem Statement

Establish quantitatively the statistics of strong vorticity fluctuations in the steady direct (enstrophy) cascade of two-dimensional Navier-Stokes turbulence, as left open by the source's own stated limitations. Concretely:

(1) Quantitative test of the tail law. Determine, by controlled high-resolution simulations and/or experiments of the forced 2D direct cascade, whether the PDF of the coarse-grained vorticity ϖ\varpi obeys ln⁡P(ϖ)≃−∣ϖ∣/[Hln⁡(L/R)]1/3\ln P(\varpi) \simeq -|\varpi|/[H\ln(L/R)]^{1/3} with the predicted dependence on the measured enstrophy flux HH and on the scale separation ln⁡(L/R)\ln(L/R) - including whether the double-logarithmic ln⁡ln⁡(L/R)\ln\ln(L/R) correction is resolvable over the accessible range - and whether the tail is universal, i.e. independent of the pumping statistics beyond the flux value. A valid answer either confirms the law quantitatively (slope consistent with the [Hln⁡(L/R)]−1/3[H\ln(L/R)]^{-1/3} scaling up to an order-unity prefactor and within the stated statistical and systematic errors, across at least two distinct pumping statistics) or establishes a definite deviation (for example, genuine intermittency corrections to the moment scaling) that survives a controlled error analysis excluding friction, finite-size, coherent-vortex, and sampling artifacts.

(2) Controlled refinement of the derivation. Place the two uncontrolled steps of the instanton derivation on controlled footing: (i) derive, rather than assume, the initial condition at the remote time t∗t_* for the second angular harmonic of the vorticity correlation - or show the assumed form is inconsistent with the equation of motion, correcting the tail accordingly; (ii) quantify the subleading corrections, namely the neglected cubic interactions of the nonzero angular harmonics and the higher cumulants in the Gaussian integration over harmonics, and delimit the parameter domain in ϖ\varpi and ln⁡(L/R)\ln(L/R) where the leading-logarithmic saddle-point answer is valid.

Items (1) and (2) are two faces of a single objective - establishing the predicted vorticity tail law as quantitative, controlled knowledge about the 2D direct cascade - and each can settle the other: a decisive deviation found in (1) would force identification of the missing corrections in (2), while a controlled derivation in (2) predicts exactly what (1) must observe. The source's further suggestions - applying the instanton approach to cascades with fast strong fluctuations such as the 3D energy cascade, and to other temporally decelerating cascades such as the 2D inverse cascade and wave-turbulence cascades - are extensions to different physical systems and constitute follow-on directions, not part of the completion boundary above.

References

  1. G. Falkovich and V. Lebedev, 'Vorticity statistics in the direct cascade of two-dimensional turbulence,' Physical Review E 83, 045301(R) (2011). DOI: 10.1103/PhysRevE.83.045301, https://doi.org/10.1103/PhysRevE.83.045301
  2. G. Falkovich, V. Lebedev, and M. Stepanov, 'Vorticity statistics in the direct cascade of two-dimensional turbulence,' preprint arXiv:1011.3647 (2010). https://arxiv.org/abs/1011.3647 (detailed derivation, cited as ref [18] of the published Rapid Communication)
  3. G. Boffetta and S. Musacchio, 'Evidence for the double cascade scenario in two-dimensional turbulence,' Physical Review E 82, 016307 (2010). DOI: 10.1103/PhysRevE.82.016307, https://doi.org/10.1103/PhysRevE.82.016307
  4. C. Pasquero and G. Falkovich, 'Stationary spectrum of vorticity cascade in two-dimensional turbulence,' Physical Review E 65, 056305 (2002). DOI: 10.1103/PhysRevE.65.056305, https://doi.org/10.1103/PhysRevE.65.056305
  5. H. S. Tan, Y. X. Huang, and J. Meng, 'Hilbert statistics of vorticity scaling in two-dimensional turbulence,' Physics of Fluids 26, 015106 (2014). DOI: 10.1063/1.4861068, https://doi.org/10.1063/1.4861068; preprint arXiv:1401.4200, https://arxiv.org/abs/1401.4200
  6. E. A. Kuznetsov and E. V. Sereshchenko, 'Isotropization of two-dimensional hydrodynamic turbulence in the direct cascade,' JETP Letters 105, 83-88 (2017). DOI: 10.1134/S0021364017020102, https://doi.org/10.1134/S0021364017020102; preprint arXiv:1702.00738, https://arxiv.org/abs/1702.00738
  7. K. Kleineberg and R. Friedrich, 'Gaussian vortex approximation to the instanton equations of two-dimensional turbulence,' Physical Review E 87, 033007 (2013). DOI: 10.1103/PhysRevE.87.033007, https://doi.org/10.1103/PhysRevE.87.033007; preprint arXiv:1212.3810, https://arxiv.org/abs/1212.3810
  8. T. Grafke, R. Grauer, and T. Schaefer, 'The instanton method and its numerical implementation in fluid mechanics,' Journal of Physics A: Mathematical and Theoretical 48, 333001 (2015). DOI: 10.1088/1751-8113/48/33/333001, https://doi.org/10.1088/1751-8113/48/33/333001; preprint arXiv:1506.08745, https://arxiv.org/abs/1506.08745
  9. T. Schorlepp, T. Grafke, S. May, and R. Grauer, 'Spontaneous symmetry breaking for extreme vorticity and strain in the three-dimensional Navier-Stokes equations,' Philosophical Transactions of the Royal Society A 380, 20210051 (2022). DOI: 10.1098/rsta.2021.0051, https://doi.org/10.1098/rsta.2021.0051
  10. R. Friedrich, M. Vosskuhle, O. Kamps, and M. Wilczek, 'Two-point vorticity statistics in the inverse cascade of two-dimensional turbulence,' Physics of Fluids 24, 125101 (2012). DOI: 10.1063/1.4767465, https://doi.org/10.1063/1.4767465
  11. T. Schorlepp, T. Grafke, and R. Grauer, 'Symmetries and zero modes in sample path large deviations,' Journal of Statistical Physics 190, 50 (2023). DOI: 10.1007/s10955-022-03051-w, https://doi.org/10.1007/s10955-022-03051-w
  12. M. Reynoso, D. Zhigunov, and R. O. Grigoriev, 'Self-similarity and the direct (enstrophy) cascade in forced two-dimensional fluid turbulence,' Journal of Fluid Mechanics (2024). DOI: 10.1017/jfm.2024.653, https://doi.org/10.1017/jfm.2024.653
  13. Q. Huang, S. Goertz, P. Hollmann, J. Conrad, C. Rohde, and M. Oberlack, 'A data-driven approach for 2D vorticity PDF equations by a new conditional average estimation,' preprint arXiv:2604.15551 (2026). https://arxiv.org/abs/2604.15551