Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Theory of the lattice Boltzmann method: discrete effects due to advection

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper shows that the directional bias of lattice Boltzmann advection is a third-order error that can be removed by tuning relaxation rates, making linear advection in D2Q9 and D2Q13 rotationally invariant to third order in space…

desk verdict Useful extension of the authors' LBM advection-error program, but the D2Q13 tuning condition is inconsistent as printed and needs correction before the paper can be trusted. read the letter →

arxiv 2411.09314 v1 pith:A77C4T4G submitted 2024-11-14 math.NA cs.NA

Pierre Lallemand (CSRC) , François Dubois (LMO , LMSSC) , Li-shi Luo (ODU , CSRC) This is my paper · ORCID
classification math.NAcs.NA MSC 65M7576M28 PACS 02.70.Ns05.20.Dd47.11.+j
keywords LatticeBoltzmannmethodTaylorexpansionHénonparametersanomalousadvectionrotationalinvarianceD2Q9modelD2Q13advection-diffusion
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Lattice Boltzmann simulations of moving patterns can distort them depending on the angle between the flow and the lattice grid. This paper attributes that distortion to odd-order space derivatives in the equivalent partial differential equation of the scheme, and shows that the distortion can be tuned away by choosing relaxation rates appropriately. For D2Q9, setting $\sigma_3=\sigma_4$ and $\sigma_4\sigma_6=1/12$ makes the third-order advection terms independent of angle; for D2Q13, the analogous condition is $\sigma_6=\sigma_8=\sigma_4/12$. With an additional special choice the residual phase-velocity error vanishes, so linear advection becomes accurate to third order in space derivatives. Simulations of an advected Gaussian dot and vortex confirm the rotational symmetry predicted by the tuned parameters.

What carries the argument

The argument is carried by the Taylor-expansion method for lattice Boltzmann schemes, which iteratively eliminates time derivatives and produces an equivalent partial differential equation whose successive space-derivative terms expose the discrete errors. Working in moment space with orthogonal polynomial moments, the authors express relaxation rates through parameters $\sigma_i=1/s_i-1/2$, and use a plane-wave dispersion analysis to isolate the third-order advection correction. The identity that matters is the isotropy condition for the third-order matrix $N_3$: once the $\theta$-dependent coefficients vanish, the model's advection is rotationally invariant to third order.

What would settle it

Simulate a Gaussian vortex in D2Q9 with the tuned rates $\sigma_3=\sigma_4$ and $\sigma_4\sigma_6=1/12$, at a finite Mach number such as 0.1, along several lattice directions, and compare the vorticity after many time steps; residual angle-dependent distortion that grows with the initial amplitude would show that the linear tuning does not control nonlinear advection errors. Equivalently, measure the phase velocity of a finite-amplitude plane wave as a function of angle and amplitude and check whether the $\sin 4\theta$ term stays zero outside the linear regime.

Watch

Extended reading notes

Core claim

The central claim is that anomalous advection in lattice Boltzmann schemes is a third-order discrete effect, expressible as an orientation-dependent correction to the phase velocity of plane waves, and that this correction can be eliminated algebraically rather than by enlarging the velocity set. In the linearized analysis around a uniform velocity, the correction takes the form $A(\theta) k^2$, where $\theta$ is the angle between the wave vector and the grid. The paper derives the exact dependence of $A$ on the relaxation parameters for D2Q9 and D2Q13, and shows that the conditions $\sigma_3=\sigma_4$ with $\sigma_4\sigma_6=1/12$ (D2Q9) or $\sigma_6=\sigma_8=\sigma_4/12$ (D2Q13) make $A$ independent of $\theta$. With $\sigma_4=1/\sqrt{12}$ in D2Q13 the remaining phase velocity vanishes, giving a model whose linear advection is rotationally invariant and accurate to third order in space derivatives; the same method yields analogous cancellation conditions for D3Q15 and D3Q19 advection-diffusion models.

Load-bearing premise

The entire parameter prescription is derived from a linearized analysis around a uniform velocity, assuming slowly varying smooth fields, and the paper's own conclusion states that only linearized situations were treated; the extension to actual nonlinear vortex flows rests on the unverified assumption that nonlinear odd-order advection errors either vanish or obey the same cancellation conditions.

Editorial extensions

If this is right

  • In D2Q9, the choices $\sigma_3=\sigma_4$ and $\sigma_4\sigma_6=1/12$ eliminate the angle dependence of the third-order advection matrix, so a linearly advected Gaussian vortex keeps its rotational symmetry.
  • In D2Q13, $\sigma_6=\sigma_8=\sigma_4/12$ removes the angular dependence of the additional phase velocity, and $\sigma_4=1/\sqrt{12}$ removes the residual phase velocity, yielding third-order accuracy.
  • For three-dimensional advection-diffusion, D3Q15 and D3Q19 admit two parameter families that suppress anomalous advection, and in the two-relaxation-times case both models share the condition $\sigma_1=1/\sqrt{12}$.
  • The analysis provides explicit formulas for the anomalous-advection factor $g(k)=k\cdot V(1+h k^2)$, allowing numerical predictions of the advection error for arbitrary wave-vector orientation, as verified in the paper's plane-wave simulations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same tuning conditions should carry over to nonlinear flows only if the neglected nonlinear odd-order terms are small compared with the linear ones; repeating the vortex simulations at higher Mach number would test this directly.
  • Because the anomalous-advection correction scales as $k^2$ of the wave vector, the rotational distortion is largest for small-scale features, so the tuning matters most in under-resolved or high-wave-number simulations.
  • The Appendix result that no single choice of rates simultaneously removes anomalous advection and hyper-diffusivity in the 3D TRT diffusion models suggests a practical trade-off: a user must decide whether to prioritize isotropic advection or fourth-order dissipation.
  • One could apply the same moment-space expansion to other lattice Boltzmann variants, such as D3Q27 or thermal models, to derive their own third-order advection isotropy conditions.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper analyzes discrete, orientation-dependent advection errors in lattice Boltzmann schemes, focusing on odd-order space derivatives. Using the so-called Berlin algorithm for Taylor expansion of the LBE in moment space, it derives equivalent equations up to third order in space derivatives for advection-diffusion (D2Q5, D3Q15, D3Q19) and athermal fluids (D2Q9, D2Q13), and proposes relaxation-rate conditions, expressed in Hénon parameters, that remove the angular dependence of the advection phase velocity. For D2Q9 athermal fluids the proposed conditions are Eq. (35), σ3 = σ4 and σ4σ6 = 1/12; for D2Q13 the paper proposes Eq. (47) and then states the resulting phase velocity Eq. (48). The paper also presents plane-wave dispersion comparisons and Gaussian vortex simulations. I verified that the D2Q9 condition (35) does make the third-order matrix isotropic as claimed, but the D2Q13 condition as printed is internally inconsistent; the correct condition is the reciprocal form σ6 = σ8 = 1/(12σ4), as explained below.

Significance. If the D2Q13 condition is corrected, the paper is a useful contribution: it gives explicit, analytic Hénon-parameter conditions that remove orientation-dependent third-order advection errors, rather than fitting parameters to simulations, and it supports the linear analysis with plane-wave simulations agreeing to within 0.15 percent. The D2Q9 result (35)-(37) is verified analytically and is a genuine strength. The main weaknesses are that the D2Q13 central condition is misprinted as written, the third-order matrices and final formulas are asserted without derivation, and the claimed third-order accuracy is established only in the linearized regime; the finite-amplitude simulations are qualitative consistency checks. These issues are fixable, but the printed central claim for D2Q13 is currently false.

major comments (3)
  1. [Section 5, 'Athermal fluid simulated with D2Q13', Eq. (47)] Equation (47) as printed, σ6 = σ8 = (1/12)σ4, does not cancel the angular terms in the D2Q13 dispersion relations. Substituting this condition into Eq. (44) leaves vφ = [10055(σ4^2 - 1)/157080] V sin(4θ), and substituting it into Eq. (46) leaves a nonzero f2(θ) term plus a further nonzero velocity-dependent term; Eq. (48) does not follow. The condition that cancels the sin(4θ) term in Eq. (44), removes the f2(θ) term in Eq. (46), and reproduces Eq. (48) is σ6 = σ8 = 1/(12σ4). This is obtained directly by solving Eq. (44) with σ6 = σ8 = σ, which gives 120660 σ4 σ = 10055, i.e., σ4 σ = 1/12. As printed, the central D2Q13 isotropy claim is false, so Eq. (47) must be corrected and the plane-wave table in Section 6 and Figure 4 re-examined under the intended condition.
  2. [Section 5, Eqs. (30)-(31) and (44)-(46)] The third-order matrix N3 for D2Q9 and the D2Q13 phase-velocity formulas are stated as results of the Berlin algorithm without derivation; the text says only that 'higher orders Ml are cumbersome and not given here.' Given the inconsistency in Eq. (47), the reader cannot check whether the stated formulas are correct without redoing the entire expansion. Please include the explicit N3 matrices for both D2Q9 and D2Q13, or provide a supplementary derivation or reproducible code that generates these formulas.
  3. [Abstract and Section 6] The abstract states that the paper proposes situations 'accurate to third order in space derivatives,' but Sections 4 and 5 analyze only linearized plane waves about a uniform velocity, as the conclusion itself acknowledges. The Gaussian vortex simulations in Figures 3 and 4 are finite-amplitude and are interpreted qualitatively, without an error measure tied to the third-order analysis. The third-order accuracy claim should be explicitly qualified as holding in the linearized regime, and the simulations should be described as consistency checks rather than validations of third-order accuracy for nonlinear flows.
minor comments (5)
  1. [Introduction and throughout] There are numerous typographical errors, e.g., 'generationg' on page 2, 'extention', 'usefull', 'ancellation', and 'basedof' in the appendices; the manuscript would benefit from a careful proofreading pass.
  2. [Section 6, Figures 2 and 4] The domain sizes '1012' and '3632' should likely read '101×101' and '363×363'; please clarify the notation.
  3. [Section 6, plane-wave table] The table for D2Q13 plane-wave advection does not list the relaxation parameters (σ4, σ6, σ8, c1, q) used in the simulation. Please provide these values so the reader can verify which isotropy condition was actually applied.
  4. [Figure 1 and Eq. (49)] The caption of Figure 1 does not define the plotted quantity h nor identify which curve is solid and which is dashed; please make the definition and legend explicit.
  5. [Appendix 2, Eq. (66) and Table 7] 'relaxation rates5' should be 'relaxation rate s5', and the table caption should state which moments correspond to the non-conserved relaxation rates s5, s6, s11, s14, s16, and s17.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the advection-error isotropy conditions are derived analytically from the moment system and only consistency-checked by simulation.

full rationale

The paper's central results are the relaxation-rate conditions (35) and (47) that make the third-order advection terms orientation-independent. These conditions are obtained by an explicit Taylor expansion of the LBE in the moment basis: for D2Q9, Eq. (31) defines the theta-dependent coefficients h1, h3, g1, g2 and g3, and Eq. (35) is precisely the system that makes those coefficients vanish, leaving the isotropic N3 of Eq. (36). For D2Q13, the angular terms in Eqs. (44) and (46) are killed by the reciprocal relation sigma6 = sigma8 = 1/(12 sigma4); the printed expression should be read as that reciprocal condition, which is consistent with Eq. (48). Nothing in this chain is fitted to the simulations shown later: the simulations compare against either the analytic dispersion prediction or the exact Gaussian solution (51)-(54), so they are consistency checks, not sources of the conditions. The Taylor-expansion algorithm is attributed to the authors' earlier papers [1]-[4], but the present manuscript displays the expanded matrices and the resulting cancellation conditions itself; the prior work supplies a general expansion procedure, not the target isotropy conclusion. The closing caveat that the analysis is linearized is a scope restriction on the nonlinear claim, not a circularity. No step reduces a prediction to its own input.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities; 'anomalous advection' is a descriptive term for the odd-order derivative correction. The central claim depends on the chosen equilibrium moment forms and relaxation rates, which are adjustable model parameters rather than fitted constants. No external physical data are used, so the ledger is dominated by modeling assumptions inherited from the authors' prior framework.

free parameters (7)
  • alpha = not fixed; examples use alpha = -2
    Equilibrium energy constant; sets sound speed and affects diffusivity; chosen by hand rather than fitted to data.
  • beta = not fixed
    Equilibrium value of a ghost moment; appears in hyper-diffusivity conditions.
  • q = -7/6 (recommended for D2Q13)
    Tuning parameter in D2Q13 equilibrium moments that controls the velocity dependence of shear viscosity.
  • c1 = not fixed
    D2Q13 heat-flux equilibrium parameter; appears in viscosity formulas and advection phase velocity expressions.
  • d1 = -1 for D2Q9 diffusion; -7/3 for D3Q15; -2/3 for D3Q19
    Diffusive equilibrium parameter; special values are used to make anomalous advection vanish.
  • d2 = 0 for D3Q19
    Extra equilibrium parameter in D3Q19; set to zero in the first isotropy condition.
  • sigma_i relaxation rates = conditions such as sigma4 sigma6 = 1/12 and sigma1 = 1/sqrt(12)
    Hénon relaxation parameters; the central claim consists of conditions on these values.
assumptions (4)
  • domain assumption The Taylor expansion method ('Berlin algorithm') from Augier et al. 2014 correctly produces equivalent PDEs to third order.
    Invoked in Section 3; the paper does not re-derive the algorithm.
  • domain assumption The moment equilibria are polynomial functions of velocity of the given forms (Tables 3, 4, 5, 6, 7).
    The derived tuning conditions depend on these chosen equilibrium forms.
  • domain assumption Solutions are smoothly varying and perturbations are small enough to linearize around a uniform velocity.
    Section 4 uses linear analysis; the conclusion admits the extension to nonlinear flows is not established.
  • ad hoc to paper The third-order odd-derivative terms dominate the advection error and can be canceled independently of even-order terms.
    The paper focuses on odd-order derivatives but does not prove that higher-order or cross terms cannot spoil rotational invariance in general.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Theory of the lattice Boltzmann method: discrete effects due to advection." pith.science (2026). https://pith.science/paper/A77C4T4G

@misc{pith2026241109314,
  author       = {Pith},
  title        = {Pith review of: Theory of the lattice Boltzmann method: discrete effects due to advection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A77C4T4G}},
  note         = {Machine review of arXiv:2411.09314}
}
read the original abstract

Lattice Boltzmann models are briefly introduced together with references to methods used to predict their ability for simulations of systems described by partial differential equations that are first order in time and low order in space derivatives. Several previous works have been devoted to analyzing the accuracy of these models with special emphasis on deviations from pure Newtonian viscous behaviour, related to higher order space derivatives of even order. The presentcontribution concentrates on possible inaccuracies of the advection behaviour linked to space derivatives of odd order. Detailed properties of advection-diffusion and athermal fluids are presented for two-dimensional situations allowing to propose situations that are accurate to third order in space derivatives. Simulations of the advection of a gaussian dot or vortex are presented. Similar results are discussed in appendices for three-dimensional advection-diffusion.

Figures

Figures reproduced from arXiv: 2411.09314 by the authors.

Figure 1
Figure 1. Advection factor for main velocity along [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Advection of an initial Gaussian disturbance simulated with diffusive D2Q9 under [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. Simulation with D2Q9. Vorticity of the velocity field from an initial gaussian [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Simulation with D2Q13. Vorticity of the velocity field from an initial Gaussian [PITH_FULL_IMAGE:figures/full_fig_p019_4.png]
Figure 5
Figure 5. Figure 5: Vorticity of the vortex with main velocity at [PITH_FULL_IMAGE:figures/full_fig_p020_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Isotropy and Galilean invariance of Lattice Boltzmann Method: Theoretical and numerical analysis using oblique dipole benchmark *

    math.AP 2026-07 conditional novelty 5.0 of 10

    D2Q9 LBM cannot recover fully Galilean-invariant second-order Navier–Stokes without controlling cubic parasitic terms; a new oblique dipole benchmark shows that Dubois heat-flux equilibria plus Augier-type relaxation ...

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages · cited by 1 Pith paper

  1. [1]

    Augier, F

    A. Augier, F. Dubois, B. Graille and P. Lallemand. ``On rotational invariance of Lattice Boltzmann schemes'', Computers and Mathematics with Applications , vol. 67 , p 239-255, 2014

    work page 2014

  2. [2]

    F. Dubois. ``Equivalent partial differential equations of a Boltzmann scheme'', Computers and mathematics with applications , vol. 55, p. 1441-1449, 2008

    work page 2008

  3. [3]

    Dubois, ``Nonlinear fourth order Taylor expansion of lattice Boltzmann schemes'', Asymptotic Analysis , vol

    F. Dubois, ``Nonlinear fourth order Taylor expansion of lattice Boltzmann schemes'', Asymptotic Analysis , vol. 127, p. 297-337, 2022

    work page 2022

  4. [4]

    Dubois, P

    F. Dubois, P. Lallemand. ``Towards higher order lattice Boltzmann schemes'', Journal of Statistical Mechanics: Theory and Experiment , P06006 doi: 10.1088/1742-5468/2009/06/P06006, 2009

    work page doi:10.1088/1742-5468/2009/06/p06006 2009

  5. [5]

    Dubois, P

    F. Dubois, P. Lallemand. ``On Triangular Lattice Boltzmann Schemes for Scalar Problems'', Communications in Computational Physics , vol. 13, p. 649-670, 2013

    work page 2013

  6. [6]

    M. H\'enon. ``Viscosity of a Lattice Gas'', Complex Systems , vol. 1 , p. 763-789, 1987

    work page 1987

  7. [7]

    d'Humi\`eres

    D. d'Humi\`eres. ``Generalized Lattice-Boltzmann Equations'', in Rarefied Gas Dynamics: Theory and Simulations , vol. 159 of AIAA Progress in Astronautics and Astronautics , p. 450-458, 1992

    work page 1992

  8. [8]

    M. Junk, A. Klar, L.S. Luo. ``Asymptotic analysis of the lattice Boltzmann equation'', Journal of Computational Physics , vol. 210 , p. 676-704, 2005

    work page 2005

Show all 12 references
  1. [9]

    Lallemand, L-S

    P. Lallemand, L-S. Luo. ``Theory of the lattice Boltzmann method: Dispersion, dissipation, isotropy, Galilean invariance, and stability'', Physical Review E , vol. 61 , p. 6546-6562, June 2000

    2000

  2. [10]

    Landau, E.M

    L.D. Landau, E.M. Lifshitz. Fluid Mechanics (Volume 6 of A Course of Theoretical Physics), Pergamon Press, 1959

    1959

  3. [11]

    Frisch, private communication (May 1985)

    U. Frisch, private communication (May 1985)

    1985

  4. [12]

    ,# (7),01444 '9=82<.342C 2! !22222222222222222222222222222222222222222222222222 z

    anomalous-advection-section-2-aout2022.tex0000644000000000000000000002601614715343203017332 0ustar rootroot 2) A brief description of the lattice Boltzmann equation sec:LBM The lattice Boltzmann equation (LBE) evolves on a d dimensional lattice r Z _d with lattice spacing r an...

    2022

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.