Pith. sign in

REVIEW 2 major objections 3 minor 3 cited by

Stability of the Active Flux Method in the Framework of Summation-by-Parts Operators

T0 review · 2 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper proves that the upwind Active Flux method is energy-stable for one-dimensional linear advection with periodic boundary conditions, using a degenerate summation-by-parts operator whose mass matrix has kernel exactly the constants.

desk verdict First SBP treatment of Active Flux, with a genuine but repairable gap in the upwind proof; worth refereeing after a minor revision. read the letter →

arxiv 2507.11068 v2 pith:DLZ5X446 submitted 2025-07-15 math.NA cs.NA

Wasilij Barsukow , Christian Klingenberg , Lisa Lechner , Jan Nordström , Sigrun Ortleb , Hendrik Ranocha This is my paper
classification math.NAcs.NA MSC 65M0665M2065M70
keywords ActiveFluxmethodsummation-by-partsoperatorsenergystabilitylinearadvectionconservationlawsupwindfinitedifferencesnullspaceconsistencyvolumemethods
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 sets out to prove that the Active Flux method, a hybrid finite-volume/finite-difference scheme for conservation laws that stores both cell averages and interface point values, is energy stable on one-dimensional periodic domains for linear advection. It succeeds by rewriting the scheme as a summation-by-parts (SBP) discretization: for the central point update a diagonal mass matrix makes the derivative operator skew-symmetric, and for the standard upwind update a degenerate, positive-semidefinite mass matrix makes the two upwind operators adjoints and introduces dissipation. Because the degenerate matrix has as its kernel exactly the constant states, the zero-energy part of any solution is stationary, and the remaining part has non-increasing energy. If the proof is right, it is the first SBP-based stability argument for Active Flux and gives a template for boundary treatments and extensions.

What carries the argument

The machinery is the degenerate upwind summation-by-parts operator built from the derivative matrices $D_-$ and $D_+$ in (3.17) and the symmetric, positive-semidefinite mass matrix $M$ in (3.18). The defining identities are $M D_+ + D_-^T M = 0$ (mutual adjointness) and $M(D_+ - D_-)$ negative semidefinite, which together convert the usual SBP energy estimate into dissipation. The delicate point is that $M$ is only semidefinite, with kernel exactly $\operatorname{span}\{1\}$; the proof therefore splits the solution into a constant kernel component, which is stationary because $D_- 1 = 0$, and a perpendicular component whose energy is non-increasing. For the central variant, the same construction gives a positive-definite diagonal $M$ (3.4), reducing to classical SBP skew-symmetry.

What would settle it

For a fixed number of cells $n$, compute the eigenvalues of the matrix $M(D_+ - D_-)$ shown in (3.23) and compare with (3.24): the theorem predicts that for every wavenumber $\theta = 2\pi k/n$ the nonzero eigenvalues are $-2(18+17\cos\theta+\cos 2\theta)/3$, which are all negative. If for any $n$ and $k$ an eigenvalue is positive, or if a numerical integration of the semi-discrete upwind scheme with periodic data ever shows $\|u\|_M^2$ increasing, the central stability claim fails.

Watch

Extended reading notes

Core claim

The central claim is Theorem 4.2: the upwind Active Flux semi-discretization $\frac{\mathrm{d}}{\mathrm{d}t}u + D_- u = 0$, formed from the cell-average update (3.14) and the upwind point update (2.9), is stable for the linear advection equation $\partial_t u + \partial_x u = 0$ with periodic boundary conditions. The proof works even though the mass matrix $M$ in (3.18) is only positive semidefinite: its kernel is exactly $\operatorname{span}\{1\}$, so the state splits into a constant kernel component, which is stationary because $D_- 1 = 0$, and a perpendicular component whose energy $\|u\|_M^2 = u^T M u$ does not increase. For the central point update (2.11), the same framework gives a classical SBP operator with the positive-definite diagonal mass matrix (3.4), and Corollary 4.1 records its stability. The paper also proves that the upwind Active Flux difference operators are nullspace-consistent, while the central version is not, and an appendix connects the SBP energy argument to von Neumann stability for this problem.

Load-bearing premise

The proof's load-bearing premise is that, on the periodic uniform grid with the exact upwind stencil, the zero-energy states are exactly the constant vectors and constants stay constant; if the grid is nonuniform, the boundary is non-periodic, or the point-update stencil is changed, this premise must be rechecked and is not automatic.

Editorial extensions

If this is right

  • The upwind Active Flux semi-discretization (4.1) with (3.14) and (2.9) is stable for $\partial_t u + \partial_x u = 0$ on a periodic uniform grid: the discrete energy $u^T M u$ never increases.
  • With the central point update (2.11), the same Active Flux framework fits a classical SBP operator with diagonal norm (3.4), so its discrete energy is conserved exactly.
  • The upwind difference operators are nullspace-consistent: $D_\pm u = 0$ if and only if $u$ is constant, so spurious stationary modes are excluded.
  • The central version is not nullspace-consistent; its nullspace also contains the alternating vector $(1,-1,1,-1,\ldots)^T$, meaning a checkerboard mode is stationary.
  • The mass matrix (3.18) of the upwind scheme can be interpreted as a chained trapezoidal rule, and the same SBP identities imply $M(D_+ - D_-)$ is dissipative; numerical experiments confirm the predicted energy behavior.

Reading between the lines

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

  • The degenerate-SBP proof is tied to periodic boundaries; for walls or inflow/outflow boundaries, one would need to add boundary terms, for example simultaneous approximation terms, and verify the kernel structure again.
  • The family of mass matrices in Lemma 3.3 contains free parameters; choosing them may change the discrete energy norm and hence the observed dissipation, an effect not explored numerically.
  • Nullspace consistency of the upwind scheme suggests it avoids the odd-even decoupling that the central scheme's checkerboard null vector permits; this could be checked by running both schemes on a grid-aligned high-frequency initial mode.
  • A nonlinear extension would require a flux-split or entropy-stable formulation, because the present proof uses only linear advection and a quadratic energy.
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 / 3 minor

Summary. The paper rewrites the semi-discrete Active Flux method for one-dimensional linear advection with periodic boundary conditions into a matrix form using cell averages and interface point values as independent unknowns. For the central point update (2.11)–(2.12), it constructs a diagonal mass matrix (3.4) and a parameterized family (3.6) making the derivative matrix (3.3) skew-symmetric with respect to the mass matrix, hence an SBP operator. For the upwind point update (2.9), it constructs the one-parameter positive semidefinite pentadiagonal mass matrix (3.18) with kernel span{1}, verifies the upwind SBP properties and the negative semidefiniteness of the dissipation operator (Lemma 3.11), and derives an energy inequality. The paper also presents a numerical experiment, proves nullspace consistency of the upwind operators, and establishes a connection to von Neumann stability in Appendix B.

Significance. The result is presented as the first SBP-based energy-stability proof for a semi-discrete Active Flux method, and that claim is plausible: the construction of explicit mass matrices and the block-circulant eigenvalue computations are careful and verifiable, and the reproducibility repository is a concrete strength. The scope is modest but honest: scalar linear advection, one dimension, periodic boundary conditions, and semi-discrete in time. The principal theorem, however, contains a genuine gap in the handling of the semidefinite mass-matrix kernel; the gap is local and repairable with a short mean-evolution argument, so the central claim appears defensible after revision.

major comments (2)
  1. [Section 4, Theorem 4.2, Eq. (4.4)] The proof does not justify the assertion that the kernel component u0 is constant. If u0(t)=c(t)1 is the Euclidean projection of u(t) onto span{1}, then c'(t) = -(1^T D_- u(t))/(2n), and using D_- as given in (3.17) one obtains 1^T D_- w = (6/Δx)(Σ_i w_{i-1/2} - Σ_i w_i), which is not zero in general. Therefore Eq. (4.4) does not follow from D_- 1 = 0, and inequality (4.3) bounds only ||u_perp||_M, not c(t). The gap is repairable: from (3.14) the mean cell average A=Σ_i u_i/n is conserved, while the mean point value P=Σ_i u_{i+1/2}/n satisfies dP/dt = (6/Δx)(A-P), so P converges to A and c(t)=(A+P)/2 remains bounded. This mean-evolution argument should be added to the proof.
  2. [Section 4, Theorem 4.2] The statement that the method is stable is incomplete because no norm is specified for the stability assertion. Since M in (3.18) is positive semidefinite, ||u||_M is a seminorm that vanishes on span{1}; the energy inequality alone therefore does not provide a bound on the full state vector. After the kernel-component control from the previous comment is added, the authors should either prove a bound in an explicit norm (e.g., the Euclidean norm or the M-seminorm combined with the conserved cell-average mean) or state the theorem as an energy estimate together with a separate boundedness statement for the kernel mode.
minor comments (3)
  1. [Section 7] There are two typographical errors in the summary paragraph: 'stabiliy' and 'von Neumann stabiliy' should read 'stability'.
  2. [Figure 2] The vertical-axis label 'Energy Chan ge' contains a typo, and the caption should state explicitly that the plotted quantity is the change in the M-seminorm, since for the upwind version the mass matrix is only positive semidefinite.
  3. [Section 5] The numerical experiment uses one initial condition and 50 volumes; adding a second case with a nonzero initial difference between the mean point value and the mean cell average would directly exercise the kernel-mode dynamics that the revised proof must control.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: stability results follow from explicit algebraic SBP identities; self-citations are background, and the Theorem 4.2 concern is a proof gap, not a circular reduction.

full rationale

The derivation chain is self-contained. The central stability result (Corollary 4.1) follows from the verified identity MD + D^T M = 0 with the explicit diagonal mass matrix in (3.4), and the upwind result rests on the adjointness identity MD_+ + D_-^T M = 0 together with semidefiniteness of M(D_+ - D_-), established directly in Lemmas 3.7, 3.9, and 3.11. These are algebraic verifications performed inside the paper, not fitted predictions or imported theorems. The mass matrix is chosen as a Lyapunov weight; choosing a norm to expose stability is a standard proof technique and is not circular. The self-citations to [5,6,8,9] describe the Active Flux construction and reproducibility artifacts, while the SBP framework cites external reviews [20,60]; none of these citations carries the stability argument. One caveat is a correctness gap rather than a circularity: in the proof of Theorem 4.2, the component of the solution in ker(M) is asserted to be constant from d/dt u0 = -D_- u0 = 0, but the Euclidean projection of a general solution onto span{1} evolves because 1^T D_- is nonzero. This makes the upwind proof incomplete as written, although the conclusion appears repairable by a separate bound on the mean point-average difference. That concern does not make the claimed result equivalent to its inputs, so no circular step is present.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The proof relies on the exact stencils of the Active Flux method, the periodic uniform grid, and standard SBP/energy arguments. The mass matrix parameters are analysis constructs, not fitted to data, and no new physical or mathematical entities are introduced.

free parameters (2)
  • m_v scaling of upwind mass matrix = 1 (chosen in Remark 3.10 for normalization)
    The mass matrix (3.18) is determined up to a positive scalar; the stability proof holds for any m_v > 0. m_v=1 is used in the numerical experiment.
  • m_p, m_vv parameters of central mass matrix family = m_vv=0, 2m_v/9 < m_p < 2m_v/3 for positive definiteness
    These free parameters define the family of admissible mass matrices (3.6). They are analysis tools; the numerical scheme does not depend on them.
assumptions (5)
  • domain assumption The semi-discrete Active Flux method as defined by (2.6) with point updates (2.9)-(2.12) is the object of analysis
    The paper analyzes the spatial semi-discretization; fully discrete time integration effects are not part of the stability proof.
  • domain assumption Periodic boundary conditions on a uniform grid with cell size Δx
    Used for the periodic SBP definitions and the block-circulant structure; non-periodic boundaries would require boundary terms.
  • domain assumption Linear advection equation with a=1; the upwind and central point updates are taken as given
    The result is restricted to the scalar linear advection equation with constant speed, as stated in the abstract.
  • standard math Block circulant matrix theory (Lemma A.1, Lemma A.2)
    Standard eigenvalue results for circulant matrices, cited from [15,62].
  • standard math The energy method: boundedness of a suitable discrete norm implies stability
    Classical SBP stability framework, see (2.17) and (2.21).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stability of the Active Flux Method in the Framework of Summation-by-Parts Operators." pith.science (2026). https://pith.science/paper/DLZ5X446

@misc{pith2026250711068,
  author       = {Pith},
  title        = {Pith review of: Stability of the Active Flux Method in the Framework of Summation-by-Parts Operators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DLZ5X446}},
  note         = {Machine review of arXiv:2507.11068}
}
read the original abstract

The Active Flux method is a numerical method for conservation laws using a combination of cell averages and point values as independent degrees of freedom, based on ideas from finite volumes and finite differences. This unusual mix has been shown to work well in many situations. We expand the theoretical justifications of the Active Flux method by analyzing it from the point of view of summation-by-parts (SBP) operators, which are routinely used to analyze finite difference, finite volume, and finite element schemes. We investigate in what type of setting the Active Flux method can be formulated using classical or degenerate SBP operators, yielding a first and novel approach for showing the energy stability of the Active Flux method. We present the analysis for the one-dimensional scalar linear advection equation with periodic boundary conditions on a uniform grid.

Figures

Figures reproduced from arXiv: 2507.11068 by the authors.

Figure 1
Figure 1. The point values are placed at the cell interfaces and shared by adjacent cells. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 3 Pith papers

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

  1. Computing Radially-Symmetric Solutions of the Ultra-Relativistic Euler Equations with Entropy-Stable Discontinuous Galerkin Methods

    math.NA 2025-08 accept novelty 7.0 of 10

    The authors derive an entropy-conservative two-point flux for the ultra-relativistic Euler equations, prove its consistency, and validate an entropy-stable DG scheme against 1D radial reference solutions in 2D and 3D.

  2. Convergence of entropy-conservative summation-by-parts discretizations to smooth solutions of hyperbolic conservation laws

    math.NA 2026-07 accept novelty 6.0 of 10

    Entropy-conservative diagonal-norm SBP flux-differencing schemes converge at order p to smooth solutions of general entropy-symmetrizable hyperbolic systems under periodic boundaries.

  3. GPU-Accelerated Energy-Conserving Methods for the Two-Dimensional Hyperbolized Serre-Green-Naghdi Equations

    math.NA 2026-01 accept novelty 6.0 of 10

    2D energy-conserving finite-difference SBP/split-form schemes for the hyperbolic Serre-Green-Naghdi equations are derived, proved, and GPU-accelerated in Julia.

Reference graph

Works this paper leans on

60 extracted references · 24 canonical work pages · cited by 3 Pith papers

  1. [1]

    Acombinationofresidualdistributionandtheactivefluxformulationsoranew classofschemesthatcancombineseveralwritingsofthesamehyperbolicproblem:applica- tion to the 1D Euler equations

    R.Abgrall.“Acombinationofresidualdistributionandtheactivefluxformulationsoranew classofschemesthatcancombineseveralwritingsofthesamehyperbolicproblem:applica- tion to the 1D Euler equations.” In:Communications on Applied Mathematics and Computation 5.1 (2023), pp. 370–402.doi:10.1007/s42967-021-00175-w. 20

    doi:10.1007/s42967-021-00175-w 2023

  2. [3]

    A Semi-discrete Active Flux Method for the Euler Equations on Cartesian Grids

    R. Abgrall, W. Barsukow, and C. Klingenberg. “A Semi-discrete Active Flux Method for the Euler Equations on Cartesian Grids.” In:Journal of Scientific Computing102.2 (2025), p. 36. doi:10.1007/s10915-024-02749-1

    work page doi:10.1007/s10915-024-02749-1 2025

  3. [4]

    Analysis of the SBP-SAT Stabilization for Finite Element Methods Part I: Linear problems

    R. Abgrall, J. Nordström, P. Öffner, and S. Tokareva. “Analysis of the SBP-SAT Stabilization for Finite Element Methods Part I: Linear problems.” In:Journal of Scientific Computing85.2 (2020), pp. 1–29.doi:10.1007/s10915-020-01349-z. arXiv:1912.08108 [math.NA]

    arXiv 2020

  4. [5]

    A hybrid finite element - finite volume method for conserva- tion laws

    R. Abgrall and W. Barsukow. “A hybrid finite element - finite volume method for conserva- tion laws.” In:Applied Mathematics and Computation447 (2023).doi:https://doi.org/10. 1016/j.amc.2023.127846

    work page arXiv 2023

  5. [6]

    Extensions of Active Flux to arbitrary order of accu- racy

    Abgrall, Remi and Barsukow, Wasilij. “Extensions of Active Flux to arbitrary order of accu- racy.” In:ESAIM: M2AN57.2 (2023), pp. 991–1027.doi:10.1051/m2an/2023004

    arXiv 2023

  6. [7]

    Stationarity preserving schemes for multi-dimensional linear systems

    W. Barsukow. “Stationarity preserving schemes for multi-dimensional linear systems.” In: Mathematics of Computation88 (2019), pp. 1621–1645.doi:10.1090/mcom/3394

    work page doi:10.1090/mcom/3394 2019

  7. [8]

    A generalized Active Fluxmethodofarbitrarilyhighorderintwodimensions

    W. Barsukow, P. Chandrashekar, C. Klingenberg, and L. Lechner. “A generalized Active Fluxmethodofarbitrarilyhighorderintwodimensions.”In:arXivpreprintarXiv:2502.05101 (2025)

    arXiv 2025

  8. [9]

    Stability of the Active Flux Method in the Framework of Summation- by-Parts Operators

    W. Barsukow, C. Klingenberg, L. Lechner, J. Nordström, S. Ortleb, and H. Ranocha.Re- producibility repository for "Stability of the Active Flux Method in the Framework of Summation- by-Parts Operators".https : / / github . com / ranocha / 2025 _ active _ flux _ sbp. 2025.doi: 10.5281/zenodo.15861045

    work page doi:10.5281/zenodo.15861045 2025

Show all 60 references
  1. [10]

    Julia:AFreshApproachtoNumerical Computing

    J.Bezanson,A.Edelman,S.Karpinski,andV.B.Shah.“Julia:AFreshApproachtoNumerical Computing.” In:SIAM Review59.1 (2017), pp. 65–98.doi:10.1137/141000671. arXiv:1411. 1607 [cs.MS]

    2017 doi

  2. [11]

    Entropy Stable Spectral Collocation Schemes for the Navier-Stokes Equations: Discontinuous Interfaces

    M. H. Carpenter, T. C. Fisher, E. J. Nielsen, and S. H. Frankel. “Entropy Stable Spectral Collocation Schemes for the Navier-Stokes Equations: Discontinuous Interfaces.” In:SIAM Journal on Scientific Computing36.5 (2014), B835–B867.doi:10.1137/130932193

    2014 doi

  3. [12]

    Time-Stable Boundary Conditions for Finite-Difference Schemes Solving Hyperbolic Systems: Methodology and Application to High-Order Compact Schemes

    M. H. Carpenter, D. Gottlieb, and S. Abarbanel. “Time-Stable Boundary Conditions for Finite-Difference Schemes Solving Hyperbolic Systems: Methodology and Application to High-Order Compact Schemes.” In:Journal of Computational Physics111.2 (1994), pp. 220– 236.doi:10.1006/jcph...

    1994

  4. [14]

    Makie.jl: Flexible high-performance data visualization for Julia

    S. Danisch and J. Krumbiegel. “Makie.jl: Flexible high-performance data visualization for Julia.” In:Journal of Open Source Software6.65 (2021), p. 3349.doi:10.21105/joss.03349

    2021 doi

  5. [15]

    P. J. Davis.Circulant Matrices. New York: Wiley Interscience, 1979

    1979

  6. [16]

    High-accuracy finite-difference schemes for solving elas- todynamic problems in curvilinear coordinates within multiblock approach

    L. Dovgilovich and I. Sofronov. “High-accuracy finite-difference schemes for solving elas- todynamic problems in curvilinear coordinates within multiblock approach.” In:Applied Numerical Mathematics93 (2015), pp. 176–194.doi:10.1016/j.apnum.2014.06.005

    2015 doi

  7. [17]

    Active flux schemes for systems

    T. Eymann and P. Roe. “Active flux schemes for systems.” In:20th AIAA computational fluid dynamics conferenceAIAA 2011-3840 (2011).doi:10.2514/6.2011-3840

    2011 doi

  8. [18]

    ActiveFluxSchemes

    T.A.EymannandP.L.Roe.“ActiveFluxSchemes.”In:49thAIAAAerospaceSciencesMeeting including the New Horizons Forum and Aerospace Exposition(2011).doi:10.2514/6.2011-382. 21

    2011 doi

  9. [19]

    MultidimensionalActiveFluxSchemes

    T.A.EymannandP.L.Roe.“MultidimensionalActiveFluxSchemes.”In:21stAIAACompu- tational Fluid Dynamics Conference(2013).doi:10.2514/6.2013-2940

    2013 doi

  10. [20]

    Review of summation-by-parts operators with simultaneous approximation terms for the numerical solution of partial differential equations

    D. C. D. R. Fernández, J. E. Hicken, and D. W. Zingg. “Review of summation-by-parts operators with simultaneous approximation terms for the numerical solution of partial differential equations.” In:Computers & Fluids95 (2014), pp. 171–196.doi:10 . 1016 / j . compfluid.2014.02.016

    2014

  11. [21]

    A Skew-Symmetric Discontinuous Galerkin Spectral Element Discretization and Its Relation to SBP-SAT Finite Difference Methods

    G. J. Gassner. “A Skew-Symmetric Discontinuous Galerkin Spectral Element Discretization and Its Relation to SBP-SAT Finite Difference Methods.” In:SIAM Journal on Scientific Com- puting35.3 (2013), A1233–A1253.doi:10.1137/120890144

    2013 doi

  12. [22]

    Multi-dimensionalsummation-by-parts operators for general function spaces: Theory and construction

    J.Glaubitz,S.-C.Klein,J.Nordström,andP.Öffner.“Multi-dimensionalsummation-by-parts operators for general function spaces: Theory and construction.” In:Journal of Computational Physics491 (2023).doi:10.1016/j.jcp.2023.112370

    2023

  13. [23]

    Summation-by-parts operators for general function spaces: The second derivative

    J. Glaubitz, S.-C. Klein, J. Nordström, and P. Öffner. “Summation-by-parts operators for general function spaces: The second derivative.” In:Journal of Computational Physics504 (2024).doi:10.1016/j.jcp.2024.112889

    2024

  14. [24]

    Summation-by-Parts Operators for General Func- tion Spaces

    J. Glaubitz, J. Nordström, and P. Öffner. “Summation-by-Parts Operators for General Func- tion Spaces.” In:SIAM Journal on Numerical Analysis61.2 (2023), pp. 733–754.doi:10.1137/ 22M1470141

    2023

  15. [25]

    Energy-Stable Global Radial Basis Function Meth- ods on Summation-By-Parts Form

    J. Glaubitz, J. Nordström, and P. Öffner. “Energy-Stable Global Radial Basis Function Meth- ods on Summation-By-Parts Form.” In:Journal of Scientific Computing98.1 (2024).doi:10. 1007/s10915-023-02427-8

    2024

  16. [26]

    Hairer, C

    E. Hairer, C. Lubich, and G. Wanner.Geometric Numerical Integration: Structure-Preserving AlgorithmsforOrdinaryDifferentialEquations.Vol.31.SpringerSeriesinComputationalMath- ematics. Berlin Heidelberg: Springer-Verlag, 2006.doi:10.1007/3-540-30666-8

    2006 doi

  17. [27]

    Hicken, G

    J. Hicken, G. Yan, and S. Kaur.Constructing stable, high-order finite-difference operators on point clouds over complex geometries. Sept. 2024. arXiv:2409.00809 [math.NA]

    2024 arXiv

  18. [29]

    Multidimensional Summation-By- PartsOperators:GeneralTheoryandApplicationtoSimplexElements

    J. E. Hicken, D. C. D. R. Fernández, and D. W. Zingg. “Multidimensional Summation-By- PartsOperators:GeneralTheoryandApplicationtoSimplexElements.”In:SIAMJournalon Scientific Computing38.4 (2016), A1935–A1958.doi:10.1137/15M1038360

    2016 doi

  19. [30]

    AFluxReconstructionApproachtoHigh-OrderSchemesIncludingDiscontin- uous Galerkin Methods

    H.T.Huynh.“AFluxReconstructionApproachtoHigh-OrderSchemesIncludingDiscontin- uous Galerkin Methods.” In:18th AIAA Computational Fluid Dynamics Conference. American Institute of Aeronautics and Astronautics. 2007.doi:10.2514/6.2007-4079

    2007 doi

  20. [31]

    Relaxation Runge-Kutta Methods: Conservation and Stability for Inner- Product Norms

    D. I. Ketcheson. “Relaxation Runge-Kutta Methods: Conservation and Stability for Inner- Product Norms.” In:SIAM Journal on Numerical Analysis57.6 (2019), pp. 2850–2870.doi: 10.1137/19M1263662. arXiv:1905.09847 [math.NA]

    2019 arXiv

  21. [32]

    Finite Element and Finite Difference Methods for Hyperbolic PartialDifferentialEquations

    H.-O. Kreiss and G. Scherer. “Finite Element and Finite Difference Methods for Hyperbolic PartialDifferentialEquations.”In:MathematicalAspectsofFiniteElementsinPartialDifferential Equations. Ed. by C. de Boor. New York: Academic Press, 1974, pp. 195–212

    1974

  22. [33]

    Towards the ultimate conservative difference scheme. IV. A new approach to numerical convection

    B. V. Leer. “Towards the ultimate conservative difference scheme. IV. A new approach to numerical convection.” In:Journal of Computational Physics23.3 (1977), pp. 276–299.doi: 10.1016/0021-9991(77)90095-X

    1977 doi

  23. [34]

    Entropy Production by Explicit Runge-Kutta Schemes

    C. Lozano. “Entropy Production by Explicit Runge-Kutta Schemes.” In:Journal of Scientific Computing76.1 (2018), pp. 521–565.doi:10.1007/s10915-017-0627-0. 22

    2018 doi

  24. [35]

    An efficient finite difference method for the shallow water equations

    L. Lundgren and K. Mattsson. “An efficient finite difference method for the shallow water equations.” In:Journal of Computational Physics422 (2020), p. 109784.doi:10.1016/j.jcp. 2020.109784

    2020

  25. [36]

    Diagonal-norm upwind SBP operators

    K. Mattsson. “Diagonal-norm upwind SBP operators.” In:Journal of Computational Physics 335 (2017), pp. 283–310.doi:10.1016/j.jcp.2017.01.042

    2017 doi

  26. [37]

    High-orderaccuratecomputations for unsteady aerodynamics

    K.Mattsson,M.Svärd,M.Carpenter,andJ.Nordström.“High-orderaccuratecomputations for unsteady aerodynamics.” In:Computers & Fluids36.3 (2007), pp. 636–649.doi:10.1016/ j.compfluid.2006.02.004

    2007

  27. [38]

    A stable and conservative high ordermulti-blockmethodforthecompressibleNavier-Stokesequations

    J. Nordström, J. Gong, E. van der Weide, and M. Svärd. “A stable and conservative high ordermulti-blockmethodforthecompressibleNavier-Stokesequations.”In:JournalofCom- putational Physics228.24 (2009), pp. 9020–9035.doi:10.1016/j.jcp.2009.09.005

    2009 doi

  28. [39]

    Finitevolumeapproximationsandstrictstabilityforhyperbolic problems

    J.NordströmandM.Björck.“Finitevolumeapproximationsandstrictstabilityforhyperbolic problems.” In:Applied Numerical Mathematics38.3 (2001), pp. 237–255.doi:10.1016/S0168- 9274(01)00027-7

    2001 doi

  29. [40]

    Finite volume methods, unstruc- turedmeshesandstrictstabilityforhyperbolicproblems

    J. Nordström, K. Forsberg, C. Adamsson, and P. Eliasson. “Finite volume methods, unstruc- turedmeshesandstrictstabilityforhyperbolicproblems.”In:AppliedNumericalMathematics 45.4 (2003), pp. 453–473.doi:10.1016/S0168-9274(02)00239-8

    2003 doi

  30. [41]

    Summation by parts, projections, and stability. I

    P. Olsson. “Summation by parts, projections, and stability. I.” In:Mathematics of Computation 64.211 (1995), pp. 1035–1065.doi:10.2307/2153512

    1995 doi

  31. [42]

    Summationbyparts,projections,andstability.II

    P.Olsson.“Summationbyparts,projections,andstability.II.”In:MathematicsofComputation 64.212 (1995), pp. 1473–1493.doi:10.2307/2153366

    1995 doi

  32. [43]

    A kinetic energy preserving DG scheme based on Gauss-Legendre points

    S. Ortleb. “A kinetic energy preserving DG scheme based on Gauss-Legendre points.” In: Journal of Scientifc Computing71 (2017), pp. 1135–1168.doi:10.1007/s10915-016-0334-2

    2017 doi

  33. [44]

    On the Stability of IMEX Upwind gSBP Schemes for 1D Linear Advection- Diffusion Equations

    S. Ortleb. “On the Stability of IMEX Upwind gSBP Schemes for 1D Linear Advection- Diffusion Equations.” In:Communications on Applied Mathematics and Computation(2023). doi:10.1007/s42967-023-00296-4

    2023 doi

  34. [45]

    On Strong Stability of Explicit Runge-Kutta Methods for Nonlinear Semi- bounded Operators

    H. Ranocha. “On Strong Stability of Explicit Runge-Kutta Methods for Nonlinear Semi- bounded Operators.” In:IMA Journal of Numerical Analysis41.1 (Jan. 2021), pp. 654–682.doi: 10.1093/imanum/drz070. arXiv:1811.11601 [math.NA]

    2021 arXiv

  35. [46]

    Optimized Runge-Kutta Methods with Automatic Step Size Control for Compressible Computational Fluid Dynamics

    H. Ranocha, L. Dalcin, M. Parsani, and D. I. Ketcheson. “Optimized Runge-Kutta Methods with Automatic Step Size Control for Compressible Computational Fluid Dynamics.” In: Communications on Applied Mathematics and Computation4 (Nov. 2021), pp. 1191–1228.doi: 10.1007/s42967-021...

    2021 arXiv

  36. [47]

    Energy Stability of Explicit Runge-Kutta Methods for Nonautonomous or Nonlinear Problems

    H. Ranocha and D. I. Ketcheson. “Energy Stability of Explicit Runge-Kutta Methods for Nonautonomous or Nonlinear Problems.” In:SIAM Journal on Numerical Analysis58.6 (Nov. 2020), pp. 3382–3405.doi:10.1137/19M1290346. arXiv:1909.13215 [math.NA]

    2020 arXiv

  37. [48]

    General Relaxation Methods for Initial-Value ProblemswithApplicationtoMultistepSchemes

    H. Ranocha, L. Lóczi, and D. I. Ketcheson. “General Relaxation Methods for Initial-Value ProblemswithApplicationtoMultistepSchemes.”In:NumerischeMathematik146(Oct.2020), pp. 875–906.doi:10.1007/s00211-020-01158-4. arXiv:2003.03012 [math.NA]

    2020 arXiv

  38. [49]

    A Broad Class of Conservative Numerical Methods for Dispersive Wave Equations

    H. Ranocha, D. Mitsotakis, and D. I. Ketcheson. “A Broad Class of Conservative Numerical Methods for Dispersive Wave Equations.” In:Communications in Computational Physics29 (4 2021), pp. 979–1029.doi:10.4208/cicp.OA-2020-0119

    2021 doi

  39. [50]

    𝐿2 Stability of Explicit Runge-Kutta Schemes

    H. Ranocha and P. Öffner. “𝐿2 Stability of Explicit Runge-Kutta Schemes.” In:Journal of Scientific Computing75.2 (May 2018), pp. 1040–1056.doi:10.1007/s10915-017-0595-4

    2018 doi

  40. [51]

    Summation-by-partsoperatorsforcorrectionprocedure via reconstruction

    H.Ranocha,P.Öffner,andT.Sonar.“Summation-by-partsoperatorsforcorrectionprocedure via reconstruction.” In:Journal of Computational Physics311 (Apr. 2016), pp. 299–328.doi: 10.1016/j.jcp.2016.02.009. arXiv:1511.02052 [math.NA]. 23

    2016 arXiv

  41. [52]

    RelaxationRunge-Kutta Methods: Fully-Discrete Explicit Entropy-Stable Schemes for the Compressible Euler and Navier-Stokes Equations

    H.Ranocha,M.Sayyari,L.Dalcin,M.Parsani,andD.I.Ketcheson.“RelaxationRunge-Kutta Methods: Fully-Discrete Explicit Entropy-Stable Schemes for the Compressible Euler and Navier-Stokes Equations.” In:SIAM Journal on Scientific Computing42.2 (Mar. 2020), A612– A638.doi:10.1137/19M12...

    2020 arXiv

  42. [53]

    On the robustness of high-order upwind summation-by-parts methods for nonlinear con- servation laws

    H. Ranocha, A. R. Winters, M. Schlottke-Lakemper, P. Öffner, J. Glaubitz, and G. J. Gassner. “On the robustness of high-order upwind summation-by-parts methods for nonlinear con- servation laws.” In:Journal of Computational Physics520 (2025), p. 113471.doi:https://doi. org/10....

    2025

  43. [54]

    A residual-based artificial viscosity finite difference method for scalar conservation laws

    V. Stiernström, L. Lundgren, M. Nazarov, and K. Mattsson. “A residual-based artificial viscosity finite difference method for scalar conservation laws.” In:Journal of Computational Physics430 (2021), p. 110100.doi:https://doi.org/10.1016/j.jcp.2020.110100

    2021

  44. [55]

    Summation by Parts for Finite Difference Approximations for𝑑/𝑑𝑥

    B. Strand. “Summation by Parts for Finite Difference Approximations for𝑑/𝑑𝑥.” In:Journal of Computational Physics110.1 (1994), pp. 47–67.doi:10.1006/jcph.1994.1005

    1994

  45. [56]

    StabilityofthefourthorderRunge-Kuttamethodfortime-dependent partial differential equations

    Z.SunandC.-W.Shu.“StabilityofthefourthorderRunge-Kuttamethodfortime-dependent partial differential equations.” In:Annals of Mathematical Sciences and Applications2.2 (2017), pp. 255–284.doi:10.4310/AMSA.2017.v2.n2.a3

    2017 doi

  46. [57]

    Strong Stability of Explicit Runge-Kutta Time Discretizations

    Z. Sun and C.-W. Shu. “Strong Stability of Explicit Runge-Kutta Time Discretizations.” In: SIAM Journal on Numerical Analysis57.3 (2019), pp. 1158–1182.doi:10.1137/18M122892X. arXiv:1811.10680 [math.NA]

    2019 arXiv

  47. [58]

    Steady-statecomputationsusingsummation-by- partsoperators

    M.Svärd,K.Mattsson,andJ.Nordström.“Steady-statecomputationsusingsummation-by- partsoperators.”In:JournalofScientificComputing24(2005),pp.79–95.doi:10.1007/s10915- 004-4788-2

    2005 doi

  48. [59]

    On the convergence rates of energy-stable finite-difference schemes

    M. Svärd and J. Nordström. “On the convergence rates of energy-stable finite-difference schemes.” In:Journal of Computational Physics397 (2019), p. 108819.doi:10.1016/j.jcp. 2019.07.018

    2019 doi

  49. [60]

    Review of summation-by-parts schemes for initial-boundary- value problems

    M. Svärd and J. Nordström. “Review of summation-by-parts schemes for initial-boundary- value problems.” In:Journal of Computational Physics268 (2014), pp. 17–38.doi:10.1016/j. jcp.2014.02.031

    2014 doi

  50. [61]

    From Semidiscrete to Fully Discrete: Stability of Runge-Kutta Schemes by the EnergyMethodII

    E. Tadmor. “From Semidiscrete to Fully Discrete: Stability of Runge-Kutta Schemes by the EnergyMethodII.”In:CollectedLecturesonthePreservationofStabilityunderDiscretization.Ed. by D. J. Estep and S. Tavener. Vol. 109. Proceedings in Applied Mathematics. Philadelphia: Society f...

    2002

  51. [62]

    Inverses of circulant matrices and block circulant matrices

    G. E. Trapp. “Inverses of circulant matrices and block circulant matrices.” In:Kyungpook Mathematical Journal13.1 (1973), pp. 11–20

    1973

  52. [63]

    A New Class of High-Order Energy Stable Flux Reconstruction Schemes

    P. E. Vincent, P. Castonguay, and A. Jameson. “A New Class of High-Order Energy Stable Flux Reconstruction Schemes.” In:Journal of Scientific Computing47.1 (2011), pp. 50–72.doi: 10.1007/s10915-010-9420-z. 24

    2011 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.