Pith. sign in

REVIEW 2 major objections 4 minor 31 references

Nonlinear orbital stability of stationary shock profiles for the Lax-Wendroff scheme

T0 review · 2 major / 4 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 proves that spectrally stable stationary shock profiles of the Lax-Wendroff scheme are nonlinearly orbitally stable: small perturbations converge to the shock profile carrying the perturbation's mass.

desk verdict Solid linear theory, overreaching nonlinear theorem: Theorem 2.5's ℓ∞ decay needs β≥1/3, not the stated β+σ≥5/12 range. read the letter →

arxiv 2411.13094 v1 pith:J2W5DZTV submitted 2024-11-20 math.AP cs.NAmath.NA

Jean-François Coulombel (IMT) , Grégory Faye (IMT) This is my paper · ORCID
classification math.APcs.NAmath.NA MSC 65M0665M1247B3535L6535L67
keywords hyperbolicconservationlawsshockwavesdifferenceapproximationsstabilityLax-WendroffschemespectralorbitalGreen'sfunction
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

The paper asks whether the Lax-Wendroff scheme, a second-order finite-difference method for hyperbolic conservation laws that produces spurious oscillations near shocks, can still keep stationary shock profiles stable. Its central claim is that spectral stability — the linearized scheme having no spectrum outside the unit disk except the simple eigenvalue $1$ — is sufficient for nonlinear orbital stability: every sufficiently small perturbation of the piecewise constant shock converges, at an algebraic rate, to a stationary shock profile whose mass equals the perturbation's mass. The result matters because the scheme's dispersive nature destroys the uniform $\ell^1$ estimates on which classical parabolic stability proofs rely, so the stability question genuinely seemed open. The paper closes it by a sharp pointwise study of the Green's function, organized around an activation function that describes how the perturbation's mass accumulates along the eigenvector of the neutral eigenvalue.

What carries the argument

The load-bearing object is the Lopatinskii determinant $\Delta(z)$, an explicitly computable holomorphic function that acts as a characteristic polynomial for the linearized scheme; it is assembled from the roots of the dispersion relations on the two sides of the shock. A simple zero at $z=1$ locates the neutral eigenvalue coming from the one-parameter family of profiles, and non-vanishing of $\Delta$ on the closed exterior of the spectral ellipse of (2.13) is exactly spectral stability. From $\Delta$ the paper builds the spatial Green's function, isolates its pole at $z=1$ with residue $H_j$ (the eigenvector), applies an inverse Laplace transform, and obtains the temporal Green's function whose leading term is the activation function $A(x,y)$ — a primitive of the free Green's function of the Lax-Wendroff transport equation — controlling how the mass of the initial condition accumulates along the eigenvector. The remainder terms are bounded by the functions $M_\ell,M_r,K_\ell,K_r$ encoding the oscillatory-diffusive profile of the dispersive scheme. Polynomial weights $\ell^p_\gamma$ enter because the free Green's function has the known sharp $n^{1/8}$ growth in $\ell^1$, so the weights compensate the weak instability and make the nonlinear bootstrap close.

What would settle it

Take the non-convex construction of Section 3.4 — a shock with $u_\ell=1$, $u_r=-1$, flux values $f(\pm1)=1$, and parameters $(\alpha_\ell,\alpha_r)=(1/3,-2/3)$ with a small oscillation tuning $\alpha_m\approx 8.79$ — for which the paper's formulas give $\Delta(2)=0$: running the scheme with a small perturbation should show growth rather than convergence, confirming that Assumption 1 is genuine. Alternatively, for any flux where sampling $\Delta$ on the spectral curve shows no extra zeros, the theory predicts the exact algebraic decay rates $n^{-\sigma}$ and $n^{-\sigma-11/24}$; a numerical run deviating from these exponents would break the central claim.

Watch

Extended reading notes

Core claim

Under the Rankine-Hugoniot relation, the Lax entropy inequalities $f'(u_r)<0<f'(u_\ell)$, the CFL restriction $\max(\alpha_\ell,|\alpha_r|)<1$, and Assumption 1 on the zeros of the Lopatinskii determinant, the paper proves that the family of stationary discrete shock profiles $\{v^\theta\}$ is nonlinearly orbitally stable. Concretely (Theorem 2.5), for any perturbation $h$ of the reference shock $u$ with $\|h\|_{\ell^1_\gamma}$ small enough, the numerical solution $u^n$ is well defined and satisfies $\|p_n\|_{\ell^1_\beta}\le C_0(1+n)^{-\sigma}\|h\|_{\ell^1_\gamma}$ and $\|p_n\|_{\ell^\infty_\beta}\le C_0(1+n)^{-\sigma-11/24}\|h\|_{\ell^1_\gamma}$ for $p_n=u^n-v^\theta$, with $\theta=\sum_j h_j$ and $\gamma=\sigma+\beta+1/8$. The paper also proves that every convex or concave flux satisfies Assumption 1, so for such fluxes — Burgers's equation included — every Lax entropy shock is orbitally stable; and it exhibits non-convex fluxes for which Assumption 1 fails, showing the assumption is not vacuous.

Load-bearing premise

The result rests on the premise, not proved in full generality, that the linearized scheme has no eigenvalue outside the unit disk except the neutral eigenvalue $1$ (a simple zero of the Lopatinskii determinant) — a condition the paper verifies only for convex or concave fluxes and shows can fail for non-convex ones.

Editorial extensions

If this is right

  • For any convex or concave flux satisfying the Lax entropy inequalities and the CFL condition, stationary shock profiles of the Lax-Wendroff scheme are nonlinearly orbitally stable; the Burgers-equation shocks used in the numerical experiments are covered.
  • Conservation forces the selection of the limit: the asymptotic profile is the one whose excess mass equals the total mass of the initial perturbation, so stability holds for the whole one-parameter family $\{v^\theta\}$, not for a single profile.
  • Decay is purely algebraic, at rates $n^{-\sigma}$ in $\ell^1_\beta$ and $n^{-\sigma-11/24}$ in $\ell^\infty_\beta$, because the eigenvalue $1$ is embedded in the continuous spectrum and there is no spectral gap; exponential convergence cannot be expected in this setting.
  • The zero-weight case of the linear theory recovers the sharp $n^{1/8}$ growth of the Lax-Wendroff scheme in $\ell^1$, confirming in the shock context that uniform $\ell^1$ stability is impossible and that weighted spaces are the natural framework.
  • The constraint $\beta+\sigma\ge 5/12$ with $0\le\sigma<\beta+1/8$ quantifies exactly which polynomial weights allow the bootstrap argument to close.

Reading between the lines

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

  • The authors leave a direct test implicit: because $\Delta$ is explicitly computable, Assumption 1 can be checked numerically for any proposed flux by sampling the determinant on the spectral curve; the non-convex examples of Section 3.4 show where such a check would find failures.
  • The activation-function machinery is likely the template for stability analysis of other higher-order dispersive schemes: the embedded neutral eigenvalue, the mass-accumulation function, and polynomial weights compensating dispersive growth should transfer, with exponents changing but the mechanism persisting.
  • One may read the sharp $n^{1/8}$ growth as intrinsic to second-order accuracy, suggesting a general principle that no uniform-in-time $\ell^1$ stability theory can exist for such schemes, so weighted spaces are the natural habitat of their stability statements.
  • Appendix A's sharp bounds on the activation function supply the key estimate for a local limit theorem for dispersive finite-difference approximations of transport, a connection the paper flags as work in progress rather than develops.
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

2 major / 4 minor

Summary. The paper analyzes the spectral, linear, and nonlinear stability of the piecewise constant stationary shock profile (2.7) for the Lax-Wendroff scheme (2.5)-(2.6) for scalar conservation laws. It introduces a Lopatinskii determinant (3.3) whose zero set locates the spectrum of the linearized operator L, proves for convex or concave fluxes that the only zero in the closed exterior region is the simple eigenvalue 1 (Lemma 3.2), and formulates Assumption 1 under which L has no other spectrum in the closed exterior (Theorem 2.3). The central technical content is a detailed construction of the spatial Green's function (Proposition 3.1), a decomposition of the temporal Green's function into an activation-function contribution and remainder terms governed by free constant-coefficient Green's functions (Theorem 4.1), and resulting linear decay estimates in polynomially weighted spaces (Theorem 4.3, (4.37a)-(4.38c)). A bootstrap argument in Chapter 5 uses these estimates to prove orbital stability of the family of stationary shock profiles (Theorem 2.5) for small perturbations, with rates ||p^n||_{ℓ1_β} ≤ C0(1+n)^{-σ}||h||_{ℓ1_γ} and ||p^n||_{ℓ∞_β} ≤ C0(1+n)^{-σ-11/24}||h||_{ℓ1_γ}.

Significance. If the parameter-range issue identified below is corrected, the paper would be a significant contribution to the stability theory of high-order numerical schemes for hyperbolic conservation laws. The explicit Lopatinskii determinant analysis for convex and concave fluxes, the fully written spatial Green's function formulas, and the sharp oscillatory integral estimates in Appendix A are concrete and checkable contributions, and the paper is honest about the failure of uniform ℓ1 estimates and about the conditional nature of Assumption 1. The conditional nonlinear stability theorem is a substantial extension of prior parabolic-type results to a dispersive setting. However, the main theorem as stated is not proven on its full parameter range because the proof uses linear and nonlinear semigroup estimates with exponents that are valid only under hidden restrictions on β. This is a load-bearing flaw that requires a corrected statement or a modified proof.

major comments (2)
  1. [§5.2.2, Eq. (4.37b)] The ℓ∞_β bound for the linear term L^{n+1}p0 in the induction step is claimed as C_L(β,γ)(2+n)^{-σ-11/24}||p0||_{ℓ1_γ}. Substituting γ1=β and γ2=γ=σ+β+1/8 into Theorem 4.3(4.37b) gives the exponent γ2−γ1+min(1/3,γ1) = σ+1/8+min(1/3,β), which equals σ+11/24 only when β≥1/3. For β<1/3 the exponent is σ+β+1/8, which is strictly smaller than σ+11/24, so the claimed inequality is not a consequence of (4.37b). The theorem statement allows β<1/3, for example β=0.3 and σ=0.125 satisfy β+σ≥5/12 and σ<β+1/8, but the decay bound used in the proof fails for such parameters. The proof therefore establishes Theorem 2.5 only on the restricted range β≥1/3, or with a slower ℓ∞ rate, and the statement must be corrected accordingly.
  2. [§5.2.2, Eqs. (4.38b)-(4.38c)] In bounding the two nonlinear sums in the ℓ∞_β estimate, the proof replaces the factors min(1/4,β) and min(1/8,β−1/8) in (4.38b) and (4.38c) by 1/4 and 1/8, respectively, using the denominators (1+n−m)^{β+7/12} and (1+n−m)^{β+1/8}. These replacements require β≥1/4; for β<1/4 the correct exponents are 2β+1/3 and β+min(1/8,β−1/8), which are smaller than the ones used. This is a second hidden restriction in the stated parameter range. The more restrictive condition β≥1/3 imposed by the linear term subsumes it, but if the authors choose instead to weaken the ℓ∞ rate, the nonlinear part of the bootstrap would still require β≥1/4.
minor comments (4)
  1. [Various] The manuscript contains several typographical errors, including 'Thank's' on p. 55, 'avoir' on p. 5, 'perurbation' on p. 16, and 'encompasses' on p. 20; these should be corrected in a revision.
  2. [§3.4] The two instability examples are convincing in outline, but the phrase "can be achieved by tuning small amplitude oscillations near the origin" is informal; a precise construction of the flux f with the stated derivative values would make the examples fully reproducible.
  3. [§3.2, (3.21)] The notation for B_ε(0) and B_ε(0) in the definition of the region Z_{ε*,η*} is not visually distinct enough in print; the reader must carefully track which of the round or square balls is intended, so a clearer notation or an explicit reminder at each occurrence would help.
  4. [§5.1] The constants C1 and C2 in §5.1 are defined by long displays that are hard to parse; splitting them into separate displayed summands and explicitly recording the hypothesis on β used for each application of Lemma 5.3 would improve readability.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: spectral stability is a hypothesis, and the Green's function and decay estimates are derived, not fitted; self-citations to [9] and [7] are prior published analyses, not load-bearing reductions.

full rationale

The paper's central chain is conditional: Assumption 1 (Section 3.2) posits that the Lopatinskii determinant Delta has a simple zero at 1 and no other zero in U\{1}; this is exactly spectral stability, not the target nonlinear result. The Green's function construction (Proposition 3.1), its meromorphic decomposition (Proposition 3.3) and the linear decay estimates (Theorems 4.1-4.3) are proved in the paper from the resolvent problem and the dispersion relations, with Appendix A supplying the free Green's function bounds. The nonlinear bootstrap (Chapter 5) applies Theorem 4.3 to the Duhamel formula; it does not import the conclusion. Citations to [9] (by one author) and [7] provide the free Green's function estimates and the algebraic nonlinear estimate Lemmas 5.1-5.2; these are published proofs with stated hypotheses that do not include Theorem 2.5, so by the review rules they are independent support rather than self-citation circularity. Assumption 1 is explicitly flagged as open to verification ('Our hope is that Assumption 1 can be easily verified (or proved not to hold) in specific situations'), which is an honest limitation, not a concealed input. The skeptical exponent mismatch in §5.2.2 (for beta<1/3, Theorem 4.3(4.37b) yields n^{-sigma-beta-1/8}, not n^{-sigma-11/24}) is a correctness gap on the stated parameter range, not a circularity, and does not change the score.

Assumptions & free parameters 2 free parameters · 7 assumptions · 2 invented entities

The central claim rests on: (i) standard complex analysis and functional calculus; (ii) domain assumptions (Rankine-Hugoniot + entropy + CFL) that define the admissible shocks; (iii) Assumption 1, the genuine spectral-stability hypothesis, proven for convex/concave fluxes but left open in general; (iv) the Smyrlis existence theorem for the shock family; and (v) the sharp free Green's function estimates from [9], extended in Appendix A. No empirical fitting occurs; the two hand-chosen parameters gamma and nu2 are constrained by the bootstrap inequalities and do not weaken the derivation.

free parameters (2)
  • gamma := sigma + beta + 1/8 = sigma + beta + 1/8
    Introduced in (2.15) to define the weight of the initial perturbation space. Chosen by hand so that the linear decay rates in Theorem 4.3 and the discrete convolution Lemma 5.3 close the bootstrap; not fitted to data.
  • nu2 := 7/12 + (sigma - 7/12)_+ = 7/12 + (sigma - 7/12)_+
    Defined in (5.6) to make Lemma 5.3 applicable to the L_theta - L term in the ell-1 estimate. A technical parameter choice, not an empirical fit.
assumptions (7)
  • standard math Levy-Wiener theorem and spectral theory of convolution operators on ell-q
    Used in Section 3.5.1 to identify the spectrum of the constant-coefficient operators L_l and L_r as the ellipses (2.13).
  • standard math Inverse Laplace transform representation of the semigroup (Equation 4.2)
    Holomorphic functional calculus for L^n with contour Gamma = (1+delta) S^1; the basis for all temporal Green's function estimates.
  • domain assumption Rankine-Hugoniot relation (2.3) and Lax entropy inequalities (2.4)
    Define the admissible stationary shock (2.2); used throughout to fix signs alpha_l > 0 > alpha_r and |alpha_r|, alpha_l < 1.
  • domain assumption CFL condition max(alpha_l, |alpha_r|) < 1 (2.12)
    Assumed in all theorems; guarantees the dispersion roots kappa_l, kappa_r split into stable/unstable modes (Lemma 3.1) and the existence of shock profiles (Theorem 2.1).
  • domain assumption Assumption 1: Delta has a simple zero at 1 and no other zero in the closed exterior of the spectral curve
    Stated in Section 3.2; the spectral stability hypothesis for Theorems 2.3-2.5. Verified for convex/concave fluxes in Lemma 3.2, but left open in general; Section 3.4 exhibits non-convex fluxes where it fails.
  • domain assumption Existence of the one-parameter family v_theta of shock profiles with exponential localization (Theorem 2.1)
    Cited from Smyrlis [28]; provides the family around which orbital stability is defined and the mass parametrization used to select v_theta.
  • domain assumption Sharp bounds for the free Green's functions of L_l and L_r (Theorems A.2, A.3, Corollary A.1)
    Taken from [9] and extended in Appendix A; these oscillatory estimates drive the activation function bounds and are the quantitative backbone of Theorems 4.1-4.3. They are proven in the appendix but are extremely technical.
invented entities (2)
  • Activation functions A_l and A_r (Equation 4.5) and the 'activation function' terminology
    purpose: Describe the leading contribution to the temporal Green's function: the accumulation of the initial mass along the eigenvector of the eigenvalue 1 of the linearized scheme. Explicitly defined by contour integrals, not postulated.
    This is an explicitly constructed mathematical object with proven bounds (Appendix A), not a new physical entity. It is listed for completeness: the term is borrowed from [7, 13] and is a genuine new object for the Lax-Wendroff shock context.
  • The Lopatinskii determinant Delta (Equation 3.3)
    purpose: A scalar holomorphic function whose zeroes locate the spectrum of the linearized operator; plays the role of the Evans function for the piecewise constant shock.
    A tool, explicitly computed, with its zero set analyzed. Not a postulated entity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nonlinear orbital stability of stationary shock profiles for the Lax-Wendroff scheme." pith.science (2026). https://pith.science/paper/J2W5DZTV

@misc{pith2026241113094,
  author       = {Pith},
  title        = {Pith review of: Nonlinear orbital stability of stationary shock profiles for the Lax-Wendroff scheme},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J2W5DZTV}},
  note         = {Machine review of arXiv:2411.13094}
}
abstract

In this article we study the spectral, linear and nonlinear stability of stationary shock profile solutions to the Lax-Wendroff scheme for hyperbolic conservation laws. We first clarify the spectral stability of such solutions depending on the convexity of the flux for the underlying conservation law. The main contribution of this article is a detailed study of the so-called Green's function for the linearized numerical scheme. As evidenced on numerical simulations, the Green's function exhibits a highly oscillating behavior ahead of the leading wave before this wave reaches the shock location. One of our main results gives a quantitative description of this behavior. Because of the existence of a one-parameter family of stationary shock profiles, the linearized numerical scheme admits the eigenvalue 1 that is embedded in its continuous spectrum, which gives rise to several contributions in the Green's function. Our detailed analysis of the Green's function describes these contributions by means of a so-called activation function. For large times, the activation function describes how the mass of the initial condition accumulates along the eigenvector associated with the eigenvalue 1 of the linearized numerical scheme. We can then obtain sharp decay estimates for the linearized numerical scheme in polynomially weighted spaces, which in turn yield a nonlinear orbital stability result for spectrally stable stationary shock profiles. This nonlinear result is obtained despite the lack of uniform ${\ell}$ 1 estimates for the Green's function of the linearized numerical scheme, the lack of such estimates being linked with the dispersive nature of the numerical scheme. This dispersive feature is in sharp contrast with previous results on the orbital stability of traveling waves or discrete shock profiles for parabolic perturbations of conservation laws.

Figures

Figures reproduced from arXiv: 2411.13094 by the authors.

Figure 2.1
Figure 2.1. The characteristics on either side of the shock. is approximated, on the time-space cell [n ∆t,(n+ 1) ∆t)×[j ∆x,(j + 1) ∆x) by a constant value u n j , that is iteratively defined with respect to n according to the formula: ∀ (n, j) ∈ N × Z , un+1 j = u n j − λ [PITH_FULL_IMAGE:figures/full_fig_p012_2_1.png] view at source ↗
Figure 2.2
Figure 2.2. Discrete shock profiles for the Lax-Wendroff scheme applied to the Burgers equation. Left: the reference shock (2.2). Middle: a discrete shock profile with same end states but negative mass difference (θ < 0 in Theorem 2.1). Right: a discrete shock profile with same end states but positive mass difference (θ > 0 in Theorem 2.1). 2.3 Numerical experiments We first present some numerical computations of stationary sho… view at source ↗
Figure 2.3
Figure 2.3. The evaluation at j = 0 and j = 1 of the family of stationary discrete shock profiles v θ , θ ∈ R. We now report on some computations that illustrate the dynamics of the numerical scheme (2.5). In [PITH_FULL_IMAGE:figures/full_fig_p018_2_3.png] view at source ↗
Figures from the paper (11 more)
Figure 2.4
Figure 2.4. Figure 2.4: Evolution of a perturbation with zero mass of the reference discrete shock (2.2). First line (from left to right): the initial condition, the solution at n = 40, the solution at n = 150. Second line (from left to right): the solution at n = 400, the solution at n = 7…
Figure 2.5
Figure 2.5. Figure 2.5: Evolution of a perturbation with zero mass of the reference discrete shock (2.2). 17 [PITH_FULL_IMAGE:figures/full_fig_p019_2_5.png]
Figure 2.6
Figure 2.6. Figure 2.6: Evolution of a perturbation with positive mass of the reference discrete shock (2.2). First line (from left to right): the initial condition, the solution at n = 100, the solution at n = 200. Second line (from left to right): the solution at n = 300, the solution at …
Figure 2.7
Figure 2.7. Figure 2.7: Evolution of a perturbation with positive mass of the reference discrete shock (2.2). 18 [PITH_FULL_IMAGE:figures/full_fig_p020_2_7.png]
Figure 3.1
Figure 3.1. Figure 3.1: Locating the spectrum of the operator L . In blue: the unit circle. In black: the curve (2.13). The region O is the complement of the grey shaded area. In red: the segment [1 − α 2 ℓ − i αℓ p 1 − α 2 ℓ , 1 − α 2 ℓ + i αℓ p 1 − α 2 ℓ ] outside of which one can holomor…
Figure 3.2
Figure 3.2. Figure 3.2: The region Zε⋆,η⋆ (in red) of Corollary 3.3. In grey: the set {ζ = eτ ∈ C | τ ∈ Bε⋆ (0)}. In blue: the unit circle S 1 . • κℓ(z), κu r (z) belong to U , κr(z), κu ℓ (z) belong to D for any z ∈ Zε⋆,η⋆ and those four functions depend holomorphically on z on Zε⋆,η⋆ , • …
Figure 3.3
Figure 3.3. Figure 3.3: An example of flux f that yields spectrally unstable shock profiles. The graph of the flux is depicted in blue and the chord between ur = −1 and uℓ = +1 is depicted in red. The Rankine-Hugoniot condition (2.3) and Oleinik’s entropy condition are satisfied. with posit…
Figure 4.1
Figure 4.1. Figure 4.1: Schematic illustration of the contour Γ and its decomposition into Γout (in red) and Γin (in blue). The contour Γin can be any path joining −η − i ε to −η + i ε, which remains within Bε(0) and passes to the right of the origin. The black bullets represent the end poi…
Figure 4.2
Figure 4.2. Figure 4.2: In blue: the contour Γin used in the regime n |αr| 2 ≤ j0 ≤ n in Lemma 4.2 and its decomposition into three parts (Γ−, Γη and Γ+). In red dash: the contour Γ−η. The black bullets represent the end points of the contours. 4.3.1 Estimates of the leading order term A n …
Figure 4.3
Figure 4.3. Figure 4.3: In blue: the contour Γin and its decomposition into five parts (Γ−, Γ ω >, Γω, Γ ω < and Γ+) in the regime where n −2/3 |αr| ≤ ω ≤ 1 − |αr|. The black bullets represent the end points of the contours. and thus −j0 = −n |αr| − (j0 − n |αr|) ≤ −n |αr| − 1 p 1 − |αr|  …
Figure 4.4
Figure 4.4. Figure 4.4: In blue: the contour Γin within Bε(0) and its decomposition into eight parts (Γ−, Γ ω >, Γ v −, Γ 0 −, Γ 0 +, Γ + v , Γ ω < and Γ+) in the regime where −ω∗ < ω ≤ −n −2/3 |αr|. The two red dots correspond to the approximate saddles of the phase |ω|τ + ζ 3 τ 3 − ζ 4 τ …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 31 canonical work pages

  1. [9]

    Coulombel

    J.-F. Coulombel. The Green’s function of the Lax-Wendroff and Beam-Warming schemes. Ann. Math. Blaise Pascal , 29(2):247–294, 2022

    work page 2022

  2. [7]

    L. Coeuret. Nonlinear orbital stability of stationary discrete shock profiles for scalar conservation laws. Preprint, arXiv:2409.18930 [math.AP] (2024), 2024

    work page Pith review arXiv 2024

  3. [1]

    Benzoni-Gavage, P

    S. Benzoni-Gavage, P. Huot, and F. Rousset. Nonlinear stability of semidiscrete shock waves. SIAM J. Math. Anal. , 35(3):639–707, 2003

    work page 2003

  4. [2]

    Benzoni-Gavage and D

    S. Benzoni-Gavage and D. Serre. Multi-dimensional hyperbolic partial differential equations. First- order systems and applications . Oxford Math. Monogr. Oxford: Oxford University Press, 2007

    work page 2007

  5. [3]

    Bouche and W

    D. Bouche and W. Weens. Analyse quantitative des sch´ emas num´ eriques pour les ´ equations aux d´ eriv´ ees partielles. Les Ulis: EDP Sciences, 2024

    work page 2024

  6. [4]

    Boutin, P

    B. Boutin, P. Le Barbenchon, and N. Seguin. Stability of finite difference schemes for the hyperbolic initial boundary value problem by winding number computations. Ann. Fac. Sci. Toulouse, Math. (6), to appear

    work page

  7. [5]

    Bultelle, M

    M. Bultelle, M. Grassin, and D. Serre. Unstable Godunov discrete profiles for steady shock waves. SIAM J. Numer. Anal. , 35(6):2272–2297, 1998

    work page 1998

  8. [6]

    L. Coeuret. Linear stability of discrete shock profiles for systems of conservation laws. Preprint, arXiv:2311.02507 [math.NA] (2023), 2023

    work page Pith review arXiv 2023

Show all 31 references
  1. [8]

    J. B. Conway. A course in functional analysis . Graduate Texts in Mathematics. Springer-Verlag, 1990

    1990

  2. [10]

    Coulombel and G

    J.-F. Coulombel and G. Faye. Generalized Gaussian bounds for discrete convolution powers. Rev. Mat. Iberoam., 38(5):1553–1604, 2022

    2022

  3. [11]

    Coulombel and G

    J.-F. Coulombel and G. Faye. The local limit theorem for complex valued sequences: the parabolic case. C. R., Math., Acad. Sci. Paris , 2024

    2024

  4. [12]

    C. M. Dafermos. Hyperbolic conservation laws in continuum physics , volume 325 of Grundlehren der mathematischen Wissenschaften . Springer-Verlag, 2016

    2016

  5. [13]

    Estep, M

    D. Estep, M. Loss, and J. Rauch. Boundedness of dispersive difference schemes. Math. Comp. , 55(191):55–87, 1990. 135

    1990

  6. [14]

    Godillon

    P. Godillon. Green’s function pointwise estimates for the modified Lax-Friedrichs scheme. M2AN Math. Model. Numer. Anal. , 37(1):1–39, 2003

    2003

  7. [15]

    Harten, J

    A. Harten, J. M. Hyman, and P. D. Lax. On finite-difference approximations and entropy conditions for shocks. Comm. Pure Appl. Math. , 29(3):297–322, 1976. With an appendix by B. Keyfitz

    1976

  8. [16]

    G. W. Hedstrom. The near-stability of the Lax-Wendroff method. Numer. Math. , 7:73–77, 1965

    1965

  9. [17]

    G. W. Hedstrom. Norms of powers of absolutely convergent Fourier series. Michigan Math. J. , 13:393–416, 1966

    1966

  10. [18]

    Jennings

    G. Jennings. Discrete shocks. Commun. Pure Appl. Math. , 27:25–37, 1974

    1974

  11. [19]

    Kr¨ oner.Numerical schemes for conservation laws

    D. Kr¨ oner.Numerical schemes for conservation laws . Wiley-Teubner Series Advances in Numerical Mathematics. John Wiley & Sons, Ltd., Chichester; B. G. Teubner, Stuttgart, 1997

    1997

  12. [20]

    Lax and B

    P. Lax and B. Wendroff. Systems of conservation laws. Comm. Pure Appl. Math. , 13:217–237, 1960

    1960

  13. [21]

    R. J. LeVeque. Finite volume methods for hyperbolic problems . Cambridge Texts in Applied Mathe- matics. Cambridge University Press, Cambridge, 2002

    2002

  14. [22]

    D. J. Newman. A simple proof of Wiener’s 1 /f theorem. Proc. Amer. Math. Soc., 48:264–265, 1975

    1975

  15. [23]

    V. V. Petrov. Sums of independent random variables. Translated from the Russian by A. A. Brown , volume 82 of Ergeb. Math. Grenzgeb. Springer-Verlag, Berlin, 1975

    1975

  16. [24]

    Randles and L

    E. Randles and L. Saloff-Coste. On the convolution powers of complex functions on Z. J. Fourier Anal. Appl., 21(4):754–798, 2015

    2015

  17. [25]

    Randles and L

    E. Randles and L. Saloff-Coste. Convolution powers of complex functions on Zd. Rev. Mat. Iberoam., 33(3):1045–1121, 2017

    2017

  18. [26]

    D. Serre. Systems of conservation laws. 1 . Cambridge University Press, 1999

    1999

  19. [27]

    D. Serre. Discrete shock profiles: existence and stability. In Hyperbolic systems of balance laws , volume 1911 of Lecture Notes in Math. , pages 79–158. Springer, 2007

    1911

  20. [28]

    Y. S. Smyrlis. Existence and stability of stationary profiles of the L W scheme. Comm. Pure Appl. Math., 43(4):509–545, 1990

    1990

  21. [29]

    Thom´ ee

    V. Thom´ ee. Stability of difference schemes in the maximum-norm. J. Differential Equations , 1:273– 292, 1965

    1965

  22. [30]

    L. N. Trefethen and M. Embree. Spectra and pseudospectra. Princeton University Press, 2005. The behavior of nonnormal matrices and operators

    2005

  23. [31]

    J. X. Xin. Multidimensional stability of traveling waves in a bistable reaction-diffusion equation. I. Commun. Partial Differ. Equations , 17(11-12):1889–1899, 1992. 136

    1992

Pith tools

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