Semi-discrete Active Flux as a Petrov-Galerkin method: the case of one-dimensional and Cartesian grids

Wasilij Barsukow

REVIEW 3 major objections 4 minor 27 references

Semi-discrete Active Flux as a Petrov-Galerkin method: the case of one-dimensional and Cartesian grids

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read Semi-discrete Active Flux schemes are Petrov-Galerkin methods: discontinuous biorthogonal test functions reproduce their updates.

desk verdict A promising variational re-framing of Active Flux that deserves refereeing, but the supplied text is unreadable and the key question about point-value evolution is unanswered in the abstract. read the letter →

arxiv 2508.15017 v2 pith:N734K54A submitted 2025-08-20 math.NA cs.NA

Wasilij Barsukow This is my paper · ORCID

classification math.NAcs.NA MSC 65M6065M08

keywords ActiveFluxhyperbolicconservationlawsPetrov-GalerkindiscontinuousGalerkincontinuousbiorthogonaltestfunctionssemi-discretemethodCartesiangrids

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 Active Flux, a numerical method for hyperbolic conservation laws whose unknowns are cell averages or moments plus shared point values at cell interfaces, is not a standalone update rule. It claims that semi-discrete Active Flux methods can be obtained from a weak formulation: choose trial functions for the solution and a special set of discontinuous test functions, biorthogonal to the trial functions, and the Galerkin projection of the PDE produces exactly the Active Flux update. This puts Active Flux between discontinuous Galerkin and continuous Galerkin: the trial space carries continuous interface point values, while the test space is piecewise discontinuous. Explicit constructions are given for arbitrary order in one dimension and for the classical third-order method on two-dimensional Cartesian meshes. A sympathetic reader cares because this turns a heuristic analogy into a precise statement of what kind of method Active Flux is.

What carries the argument

The machinery is the Petrov-Galerkin weak form, a variational formulation in which trial and test spaces differ, with biorthogonal test functions. The trial basis holds the Active Flux unknowns (cell moments and interface point values); the test functions are chosen so that each pairs with one unknown and the mass matrix becomes the identity. Biorthogonality reduces the weak form to explicit ODEs for the moments, while the shared point values are evolved by a separate semi-discrete rule. The trial space is continuous at the shared points; the test space is discontinuous across cell boundaries, which is what places AF between DG and CG. In the 2D Cartesian construction the test functions must

What would settle it

Pick a stated order in one dimension, construct the claimed biorthogonal test functions, assemble the corresponding mass matrix and flux integrals, and compare the resulting semi-discrete ODE coefficient by coefficient with the Active Flux update; any mismatch, or any order at which the biorthogonality system is singular, disproves the variational derivation. The two-dimensional statement can be tested at a shared corner: if no discontinuous test function reproduces the corner point-value contribution while preserving the other cell updates, the Cartesian claim fails.

Watch

Extended reading notes

Core claim

The central claim is constructive: semi-discrete Active Flux updates are the Petrov-Galerkin projection of the conservation law against discontinuous biorthogonal test functions. Active Flux stores cell averages/moments plus shared point values at interfaces; the trial space is continuous at those points, the test space discontinuous, giving the hybrid DG/CG structure previously noted heuristically. Explicit test functions are constructed for arbitrary-order Active Flux with additional moments in 1D and for the classical third-order method on 2D Cartesian grids. If correct, AF is not an ad hoc recipe but a specific member of the Petrov-Galerkin family.

Load-bearing premise

The load-bearing premise is that discontinuous biorthogonal test functions can be constructed whose weak-form projection reproduces Active Flux exactly, for every order in 1D and at shared corners in 2D, while interface point values evolve by a separate rule outside the weak form.

Editorial extensions

If this is right

Active Flux inherits the variational framework, so stability, error, and conservation analyses used for Galerkin methods can be applied directly to it.
The explicit test-function construction gives a systematic recipe for designing arbitrarily high-order Active Flux schemes with additional moments in one dimension.
The classical third-order Active Flux on Cartesian grids is shown to fit the same framework, making the interpretation immediately testable on existing implementations.
The trial/test split sharpens the sense in which AF is intermediate between DG and CG, replacing a heuristic analogy with a precise membership statement.

Reading between the lines

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

Reader's extension: if biorthogonal test functions exist at all orders, changing the test space while keeping the trial space could generate a family of hybrid DG/CG schemes rather than a single method.
Reader's extension: the variational form invites standard Petrov-Galerkin error analysis, such as inf-sup and adjoint-consistency arguments, which could yield stability estimates not present in the original derivation.
Reader's extension: a natural next test is extending the 2D Cartesian construction to unstructured meshes, where the corner consistency condition would likely be the main obstacle and a counterexample on triangles would clarify how much depends on Cartesian structure.

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.

Desk Editor's Note

read the letter

Two things you should know. First, the abstract is genuinely promising: it claims to turn the old observation that Active Flux sits between DG and CG into a formal Petrov-Galerkin construction, with explicit biorthogonal test functions for arbitrary order in 1D and the classical third-order scheme on Cartesian 2D meshes. That is a real service to the AF community, because a variational characterization is the natural route to stability and error analysis. Second, I could not check a single equation: the supplied full text is encoding-corrupted and ends with another paper's header. So my read is from the abstract alone.

What is actually new: the formal framing itself, if the constructions exist. Earlier remarks about AF combining FV/DG and CG were heuristic; the abstract says this paper gives explicit test functions. That is a concrete, checkable claim. Good.

The soft spots. The main one is the one the stress-test note flags: AF's degrees of freedom are moments plus shared point values at interfaces. The moment equations can be made to match AF by picking biorthogonal test functions—that is, the construction is answer-informed, which is fine as a characterization. But the point-value evolution in classical AF is a characteristic/exact update, not obviously a Galerkin equation. If the paper derives that row from the same weak form, the claim 'Active Flux is obtained from a variational formulation' is fully true. If the point-value rule is bolted on externally, then the variational claim covers only the moment part and the method is a hybrid PG-plus-characteristic scheme. That distinction matters and the abstract does not resolve it. I would want a referee to pin it down before accepting the headline as stated.

Also, the 2D case is only third order, and only at shared corner values, so the generality is more limited than the 1D claim—worth noting but not a flaw if stated clearly.

Who this is for: people working on high-order methods for hyperbolic conservation laws, especially the Active Flux community. If the constructions are correct, it gives them a variational language and opens a path to error analysis.

My recommendation: this deserves a serious referee. The claim is well-posed and checkable; the math is exactly the kind a competent referee can verify. Send it to review, but first make sure you are working from a readable copy of the manuscript—the arXiv version I was sent is corrupted, and that needs to be fixed before any referee can do the job.

Referee Report

3 major / 4 minor

Summary. The abstract claims that semi-discrete Active Flux (AF) methods for hyperbolic conservation laws can be obtained from a variational formulation by choosing a particular set of discontinuous, biorthogonal test functions. The paper then positions AF as an intermediate between discontinuous Galerkin and continuous Galerkin methods. Explicit constructions are announced for arbitrarily high-order AF with additional moments in one space dimension and for the classical third-order AF on two-dimensional Cartesian meshes. The supplied full text, however, is not readable: it is an encoding-corrupted sequence of characters with no coherent equations, and it ends with an unrelated arXiv header for a different paper. As a result, the technical content cannot currently be checked.

Significance. If the claimed construction is correct, the result would be a valuable structural characterization: it shows that AF, which uses cell moments and shared interface point values, can be viewed as a Petrov-Galerkin method with known trial and test spaces. This would clarify the relation of AF to DG and CG and could open systematic paths for analysis and extension. However, because the body of the manuscript is unreadable in the submitted form, the claimed explicit constructions and the biorthogonality relations cannot be verified. The significance is therefore conditional: the idea is interesting and the claim is plausible, but the present version does not yet provide the evidence needed to substantiate it.

major comments (3)

[Full Text (entire body)] The submitted full text is not readable: it consists of corrupted character fragments and no coherent equations, derivations, or proof sketches. The line ending the material, 'arXiv:2508.15014v2 [gr-qc] 25 Mar 2026', is an unrelated identifier, not part of this manuscript. The central claim about the Petrov-Galerkin construction cannot be checked because the explicit trial space, test space, and biorthogonality relations are inaccessible. A legible resubmission is required before any technical evaluation is possible.
[Abstract] The abstract states that semi-discrete Active Flux methods 'can be obtained from a variational formulation' via a particular choice of test functions. In classical AF, the shared interface point values are advanced by an exact or characteristic-based rule that is not obviously a row of a variational semi-discrete ODE. The abstract does not state whether the point-value evolution is derived from the same weak form (e.g., through a trace functional or distributional test) or is supplied as external input. If the latter, the claim is valid only for the moment/mean equations, and the full AF scheme is a hybrid variational-plus-characteristic method rather than 'obtained from' the variational formulation. The authors must state this explicitly and, if the point-value rule is external, revise the abstract's claim accordingly.
[Abstract and title] The verb 'obtained' suggests a first-principles variational derivation. The described procedure -- choosing biorthogonal test functions so that the moment equations reproduce the known AF update rules -- is an a-posteriori characterization: it establishes that AF can be identified with a member of a Petrov-Galerkin family, not that AF follows from variational principles without prior knowledge of its update. This is a legitimate and useful contribution, but the framing should consistently say 'characterized as' or 'equivalently represented as' rather than 'obtained from', to avoid overclaiming the direction of the construction.

minor comments (4)

[Abstract] The abstract is qualitative and contains no equations. Given that the contribution is an explicit algebraic construction, the abstract should state the trial space, the test space, and the biorthogonality condition in precise notation, at least for the lowest-order case.
[Full Text (ending)] The trailing line 'arXiv:2508.15014v2 [gr-qc] 25 Mar 2026' is a header from a different preprint and should be removed; its presence indicates that the submitted PDF/text has been assembled incorrectly.
[General] The term 'semi-discrete' is used without a clear definition. The authors should state whether time is kept continuous for both the moments and the point values, or whether the point-value evolution is a separate ODE coupled only through source terms.
[General] The paper does not discuss stability, accuracy, or convergence of the resulting Petrov-Galerkin method. If these are outside the scope, the authors should say so explicitly; otherwise the reader is left wondering whether the equivalence has practical implications.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper presents an equivalence/characterization of Active Flux within a Petrov-Galerkin framework, not a prediction fitted to itself.

full rationale

The abstract's central claim is an existence/equivalence theorem: with a particular choice of discontinuous biorthogonal test functions, the semi-discrete Petrov-Galerkin formulation reproduces the Active Flux update rules. This is a classification/characterization result, not a derivation from first principles in the sense of predicting an unknown method. The test functions are not fitted parameters or data-derived quantities; they are explicitly constructed, and the claimed equivalence is independently checkable by verifying the weak form against the known AF updates. Even if the construction was guided by prior knowledge of AF, that does not make the argument circular: the target method is not assumed as a premise of the weak form; the proof must establish the equality. The separate concern about whether shared interface point values are evolved from the weak form or supplied externally is a completeness/correctness issue, not a circularity. No load-bearing self-citations appear in the abstract, and no equation in the readable portions of the provided text reduces to its own input by construction. The supplied full text is heavily encoding-corrupted and ends with an unrelated arXiv header, preventing inspection of the detailed derivations, but based on the title and abstract no circular step is identifiable.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No parameters are fitted and no entities are invented: this is a constructive theory paper. What the reader is asked to take on trust includes standard weak-form machinery, the AF structure with shared point values evolved separately, and the paper's own selection of test functions chosen to reproduce known AF update rules. All three are provisional here because the full text is unreadable.

assumptions (3)

standard math The standard weak-form / variational machinery for semi-discrete conservation laws (integration by parts, flux boundary terms) extends to discontinuous test functions and biorthogonal bases.
The Petrov-Galerkin framing implicitly relies on this machinery; the abstract invokes it via 'variational formulation' and 'biorthogonal test functions.'
domain assumption Active Flux point values are shared across interfaces and are evolved by a separate rule (exact or characteristic-based evolution) rather than determined by the variational form.
The abstract describes AF degrees of freedom as 'averages/moments and (shared) point values at cell interfaces.' The paper's variational derivation is conditional on this DOF structure, and the 'semi-discrete' qualifier in the title signals that the point-value evolution is an input.
ad hoc to paper The particular discontinuous biorthogonal test functions can be chosen so that the resulting moment equations reproduce the known AF update rules, at arbitrary order in 1D and at third order on Cartesian 2D meshes.
The abstract says the formulation follows from 'a particular choice of (biorthogonal) test functions' with 'explicit constructions.' This is the paper's own constructive claim, not an independently motivated basis, and it is the step that would fail if the claimed formulas are wrong.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Semi-discrete Active Flux as a Petrov-Galerkin method: the case of one-dimensional and Cartesian grids." pith.science (2026). https://pith.science/paper/N734K54A

@misc{pith2026250815017,
  author       = {Pith},
  title        = {Pith review of: Semi-discrete Active Flux as a Petrov-Galerkin method: the case of one-dimensional and Cartesian grids},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N734K54A}},
  note         = {Machine review of arXiv:2508.15017}
}

read the original abstract

Active Flux (AF) is a numerical method for hyperbolic conservation laws, whose degrees of freedom are averages/moments and (shared) point values at cell interfaces. It has been noted previously in a heuristic fashion that it thus combines ideas from Finite Volume/Discontinuous Galerkin (DG) methods with a continuous approximation common in continuous Finite Element (CG) methods. This work shows that semi-discrete Active Flux methods can be obtained from a variational formulation through a particular choice of (biorthogonal) test functions. These latter being discontinuous, the new formulation emphasizes the intermediate nature of AF between DG and CG. Explicit constructions are given for the case of arbitrarily high-order Active Flux with additional moments in 1-d, and for the classical third-order Active Flux on two-dimensional Cartesian meshes.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 26 canonical work pages

[1]

Extensions of A ctive F lux to arbitrary order of accuracy

Remi Abgrall and Wasilij Barsukow. Extensions of A ctive F lux to arbitrary order of accuracy. ESAIM: Mathematical Modelling and Numerical Analysis , 57(2):991--1027, 2023

work page 2023
[2]

R \'e mi Abgrall. A combination of R esidual D istribution and the A ctive F lux formulations or a new class of schemes that can combine several writings of the same hyperbolic problem: application to the 1d E uler equations. Communications on Applied Mathematics and Computation , pages 1--33, 2022

work page 2022
[3]

A semi-discrete A ctive F lux method for the E uler equations on C artesian grids

R \'e mi Abgrall, Wasilij Barsukow, and Christian Klingenberg. A semi-discrete A ctive F lux method for the E uler equations on C artesian grids. Journal of Scientific Computing , 102(2):36, 2025

work page 2025
[4]

The active flux scheme for nonlinear problems

Wasilij Barsukow. The active flux scheme for nonlinear problems. Journal of Scientific Computing , 86(1):1--34, 2021

work page 2021
[5]

A generalized A ctive F lux method of arbitrarily high order in two dimensions

Wasilij Barsukow, Praveen Chandrashekar, Christian Klingenberg, and Lisa Lechner. A generalized A ctive F lux method of arbitrarily high order in two dimensions. arXiv preprint arXiv:2502.05101 , 2025

arXiv 2025
[6]

The active flux scheme on C artesian grids and its low M ach number limit

Wasilij Barsukow, Jonathan Hohm, Christian Klingenberg, and Philip L Roe. The active flux scheme on C artesian grids and its low M ach number limit. Journal of Scientific Computing , 81(1):594--622, 2019

work page 2019
[7]

Analysis of the multi-dimensional semi-discrete Active Flux method using the Fourier transform

Wasilij Barsukow, Janina Kern, Christian Klingenberg, and Lisa Lechner. Analysis of the multi-dimensional semi-discrete A ctive F lux method using the F ourier transform. accepted; arXiv preprint arXiv:2412.03477 , 2024

work page Pith review arXiv 2024
[8]

Proving stability of the active flux method in the framework of summation-by-parts operators

Wasilij Barsukow, Christian Klingenberg, Lisa Lechner, Jan Nordstr \"o m, Sigrun Ortleb, and Hendrik Ranocha. Proving stability of the active flux method in the framework of summation-by-parts operators. submitted , 2025

work page 2025

Show all 27 references

[9]

Active F lux methods for hyperbolic systems using the method of bicharacteristics

Erik Chudzik, Christiane Helzel, and M \'a ria Luk \'a c ov \'a -Medvid’ov \'a . Active F lux methods for hyperbolic systems using the method of bicharacteristics. Journal of Scientific Computing , 99(1):16, 2024

2024
[10]

Active F lux methods for hyperbolic conservation laws— F lux V ector splitting and bound-preservation

Junming Duan, Wasilij Barsukow, and Christian Klingenberg. Active F lux methods for hyperbolic conservation laws— F lux V ector splitting and bound-preservation. SIAM Journal on Scientific Computing , 47(2):A811--A837, 2025

2025
[11]

Evaluating local approximations of the l2-orthogonal projection between non-nested finite element spaces

Thomas Dickopf and Rolf Krause. Evaluating local approximations of the l2-orthogonal projection between non-nested finite element spaces. Numerical Mathematics: Theory, Methods and Applications , 7(3):288--316, 2014

2014
[12]

Finite E lements I : A pproximation and I nterpolation , volume 72

Alexandre Ern and Jean-Luc Guermond. Finite E lements I : A pproximation and I nterpolation , volume 72. Springer Nature, 2021

2021
[13]

Multidimensional active flux schemes

Timothy A Eymann and Philip L Roe. Multidimensional active flux schemes. In 21st AIAA computational fluid dynamics conference , 2013

2013
[14]

Investigations of a new scheme for wave propagation

Doreen Fan and Philip L Roe. Investigations of a new scheme for wave propagation. In 22nd AIAA Computational Fluid Dynamics Conference , page 2449, 2015

2015
[15]

Biorthogonal refinable spline functions

Tim NT Goodman. Biorthogonal refinable spline functions. Curve and Surface Fitting: Saint-Malo 1999, Cohen, Albert and Schumaker, Larry L and Rabut, Christophe , 2000

1999
[16]

High order biorthogonal functions in (curl)

Tim Haubold, Sven Beuchler, and Joachim Sch \"o berl. High order biorthogonal functions in (curl). SIAM Journal on Numerical Analysis , 62(5):2331--2348, 2024

2024
[17]

A new ADER method inspired by the active flux method

Christiane Helzel, David Kerkmann, and Leonardo Scandurra. A new ADER method inspired by the active flux method. Journal of Scientific Computing , 80(3):1463--1497, 2019

2019
[18]

Some properties of biorthogonal polynomials

Joseph DE Konhauser. Some properties of biorthogonal polynomials. Journal of Mathematical Analysis and Applications , 11:242--260, 1965

1965
[19]

Biorthogonal bases with local support and approximation properties

Bishnu Lamichhane and Barbara Wohlmuth. Biorthogonal bases with local support and approximation properties. Mathematics of Computation , 76(257):233--249, 2007

2007
[20]

On polynomial reproduction of dual fe bases

Peter Oswald and Barbara Wohlmuth. On polynomial reproduction of dual fe bases. In Thirteenth international conference on domain decomposition methods , pages 85--96, 2001

2001
[21]

An investigation of biorthogonal polynomials derivable from ordinary differential equations of the third order

Stanley Preiser. An investigation of biorthogonal polynomials derivable from ordinary differential equations of the third order. Journal of Mathematical Analysis and Applications , 4(1):38--64, 1962

1962
[22]

Penetration and diffusion of x-rays

Lewis V Spencer and U Fano. Penetration and diffusion of x-rays. calculation of spatial distributions by polynomial expansion. Physical review , 81(3):464, 1951

1951
[23]

Finite element interpolation of nonsmooth functions satisfying boundary conditions

L Ridgway Scott and Shangyou Zhang. Finite element interpolation of nonsmooth functions satisfying boundary conditions. Mathematics of computation , 54(190):483--493, 1990

1990
[24]

Biorthogonalit \'e et th \'e orie des op \'e rateurs

Philippe Tchamitchian. Biorthogonalit \'e et th \'e orie des op \'e rateurs. Revista Matem \'a tica Iberoamericana , 3(2):163--189, 1987

1987
[25]

Towards the ultimate conservative difference scheme

Bram van Leer. Towards the ultimate conservative difference scheme. IV . A new approach to numerical convection. Journal of computational physics , 23(3):276--299, 1977

1977
[26]

A mortar finite element method using dual spaces for the lagrange multiplier

Barbara I Wohlmuth. A mortar finite element method using dual spaces for the lagrange multiplier. SIAM journal on numerical analysis , 38(3):989--1012, 2000

2000
[27]

A high-order hybrid finite difference--finite volume approach with application to inviscid compressible flow problems: a preliminary study

Xianyi Zeng. A high-order hybrid finite difference--finite volume approach with application to inviscid compressible flow problems: a preliminary study. Computers & Fluids , 98:91--110, 2014

2014

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Extensions of A ctive F lux to arbitrary order of accuracy

Remi Abgrall and Wasilij Barsukow. Extensions of A ctive F lux to arbitrary order of accuracy. ESAIM: Mathematical Modelling and Numerical Analysis , 57(2):991--1027, 2023

work page 2023

[2] [2]

R \'e mi Abgrall. A combination of R esidual D istribution and the A ctive F lux formulations or a new class of schemes that can combine several writings of the same hyperbolic problem: application to the 1d E uler equations. Communications on Applied Mathematics and Computation , pages 1--33, 2022

work page 2022

[3] [3]

A semi-discrete A ctive F lux method for the E uler equations on C artesian grids

R \'e mi Abgrall, Wasilij Barsukow, and Christian Klingenberg. A semi-discrete A ctive F lux method for the E uler equations on C artesian grids. Journal of Scientific Computing , 102(2):36, 2025

work page 2025

[4] [4]

The active flux scheme for nonlinear problems

Wasilij Barsukow. The active flux scheme for nonlinear problems. Journal of Scientific Computing , 86(1):1--34, 2021

work page 2021

[5] [5]

A generalized A ctive F lux method of arbitrarily high order in two dimensions

Wasilij Barsukow, Praveen Chandrashekar, Christian Klingenberg, and Lisa Lechner. A generalized A ctive F lux method of arbitrarily high order in two dimensions. arXiv preprint arXiv:2502.05101 , 2025

arXiv 2025

[6] [6]

The active flux scheme on C artesian grids and its low M ach number limit

Wasilij Barsukow, Jonathan Hohm, Christian Klingenberg, and Philip L Roe. The active flux scheme on C artesian grids and its low M ach number limit. Journal of Scientific Computing , 81(1):594--622, 2019

work page 2019

[7] [7]

Analysis of the multi-dimensional semi-discrete Active Flux method using the Fourier transform

Wasilij Barsukow, Janina Kern, Christian Klingenberg, and Lisa Lechner. Analysis of the multi-dimensional semi-discrete A ctive F lux method using the F ourier transform. accepted; arXiv preprint arXiv:2412.03477 , 2024

work page Pith review arXiv 2024

[8] [8]

Proving stability of the active flux method in the framework of summation-by-parts operators

Wasilij Barsukow, Christian Klingenberg, Lisa Lechner, Jan Nordstr \"o m, Sigrun Ortleb, and Hendrik Ranocha. Proving stability of the active flux method in the framework of summation-by-parts operators. submitted , 2025

work page 2025