REVIEW 3 major objections 2 minor
Numerical study on the effect of geometric approximation error in the numerical solution of PDEs using a high-order curvilinear mesh
T0 review · 3 major / 2 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read Geometric approximation error can dominate inaccuracy and instability in time-dependent PDE solvers on curved domains.
desk verdict A solid numerical study on a real error source; the causal claim needs an independent geometric-error measurement before it fully lands. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is the high-order curvilinear surface mesh with negligible geometric approximation error, generated from CAD geometry by an adapted surface-mesh generator. This mesh isolates the geometry effect: because the curved surface is represented nearly exactly at the polynomial order used, any accuracy gap between it and a traditional mesh can be attributed to geometric approximation error. The method of moving frames supplies the geometric formulation that lets time-dependent PDEs be written and solved on the sphere, and the differential-operator comparisons on a spherical element provide the accuracy probe.
What would settle it
At a fixed polynomial order $p$, run the four sphere PDE tests while refining only the number of elements of a traditional straight-sided mesh; if its accuracy and conservation errors fall to the same levels as the high-order curvilinear mesh at comparable degrees of freedom, geometric approximation error is not the controlling factor.
Extended reading notes
Core claim
The paper's central claim is that geometric approximation error, not only the usual discretization error or mesh size, can determine whether a high-order numerical scheme for time-dependent PDEs on curved domains remains accurate and conservative. The claim is supported by comparing a CAD-based high-order surface mesh with negligible geometric approximation error against traditional meshes with non-negligible geometric error, first on the accuracy of differential operators on a curved element of the sphere and then on four time-dependent PDEs on the sphere solved with the method of moving frames. The intended conclusion is that at the same polynomial order $p$, the high-order curvilinear mesh maintains accuracy and conservation where the traditional meshes degrade, and that the degradation is attributable to geometry rather than to the numerical scheme.
Load-bearing premise
The comparison assumes that the CAD-based high-order surface mesh used in the tests really has negligible geometric approximation error at the polynomial order $p$ being tested; if it does not, the observed differences cannot be attributed to geometric approximation error alone.
Editorial extensions
If this is right
- Time-dependent solvers on curved domains can keep accuracy and conservation at high polynomial order without needing very fine meshes, as long as the surface mesh tracks the true geometry.
- The sphere tests give a clean benchmark that separates geometric approximation error from discretization error, so future high-order schemes can be evaluated on geometry quality.
- Refining only element size while ignoring boundary representation may waste computational effort if geometric approximation error is the dominant error.
- Adapting CAD-based mesh generation to surfaces makes high-order time-dependent simulation on curved geometries more reliable.
Reading between the lines
- The same mechanism likely applies to other curved manifolds, since it hinges on the mesh's representation of the surface rather than on the specific sphere equations.
- A direct test would be to fix $p$ and refine only the interior node placement of a curved mesh, then check whether conservation error shrinks at the same rate as the geometric approximation error.
- If the claim holds, mesh-quality indicators for curved meshes should report how far interior nodes deviate from the true surface, not just how well vertices are placed.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies how geometric approximation error — the error in the location of vertices and interior grid points on curved boundaries or surfaces — affects the accuracy and conservation properties of high-order numerical PDE solvers. The authors propose to isolate this error by generating a high-order curvilinear surface mesh with NekMesh, which they claim has negligible geometric approximation error even for high polynomial order p, and to compare it against traditional meshes with non-negligible geometric error. Two numerical experiments are outlined: a comparison of differential-operator accuracy on a curved spherical element, and solutions of four time-dependent PDEs on the sphere via the moving-frames method, with focus on accuracy and conservation.
Significance. If the central claim is supported by the full numerical evidence, the paper would make a useful contribution to high-order mesh generation and to the understanding of geometric error in PDE discretizations on curved domains. The comparison against traditional meshes is an external benchmark rather than a circular self-consistency test, and the moving-frames formulation is independent of the geometric mesh error, which strengthens the methodological design. The explicit attention to conservation properties is a valuable addition to the literature, where such effects are often overlooked.
major comments (3)
- [Abstract] The abstract asserts that the NekMesh surface mesh has 'negligible geometric approximation error' but provides no quantitative definition, measurement, or reference to where such a measure can be found. This premise is load-bearing: the entire causal attribution of accuracy and conservation differences to geometric approximation error depends on an independent verification that the NekMesh mesh actually has negligible geometric error at the polynomial orders tested. Please state how this error is measured (for example, distance to the exact CAD surface or a geometric-error convergence study) and report the values that support the negligibility claim.
- [Abstract] The abstract does not report any numerical results: there are no convergence tables, no error norms for the differential operators, no conservation metrics for the four PDE tests, and no stability comparisons. As a consequence, the central claim that geometric approximation error causes 'inaccuracy and instability' and affects 'conservation properties' is not substantiated by the abstract alone. Since the abstract is the only text provided for review, this is a major omission. Please add quantitative summaries of the outcomes of both test types, such as error reductions or conservation-error magnitudes.
- [Abstract] The comparison between NekMesh and 'traditional meshes with non-negligible geometric approximation error' is not sufficiently controlled in the description. It is unclear whether the two mesh types differ only in the geometric location of boundary and interior points, or also in element shape regularity, quadrature rule, or the discrete moving-frames operator. Without an explicit statement that all other discretization parameters are held fixed, the observed differences cannot be uniquely attributed to geometric approximation error. Please clarify the experimental control.
minor comments (2)
- [Abstract] The sentence 'which seems to necessitate very fine meshes especially to remove geometric approximation error' is grammatically awkward and could be clarified; rephrasing would improve readability.
- [Abstract] The acronym 'p' is used for polynomial order without explicit definition; consider defining it at first use.
Circularity Check
No circularity detected: the comparative mesh-error study uses an external benchmark and no fitted parameter is renamed as a prediction.
full rationale
This is an abstract-only review, so the full derivation chain cannot be inspected, but nothing in the available text exhibits a circular step. The paper's claim is comparative: it contrasts a high-order curvilinear surface mesh generated by NekMesh (described as having negligible geometric approximation error) with traditional meshes (described as having non-negligible geometric approximation error), and it assesses accuracy of differential operators and conservation properties of four PDEs on the sphere. The comparison against traditional meshes is an external benchmark, not an internal fit. No parameter is fitted to a subset of data and then 'predicted' on a closely related quantity. No uniqueness theorem or load-bearing self-citation is invoked. The moving-frames formulation is an independent numerical apparatus used to solve the PDEs, and the geometric error attribution is tested by comparing mesh types rather than assumed by construction. The abstract does not quantify the geometric approximation error of the NekMesh mesh, which is a weakness in supporting the causal attribution, but that is a correctness or evidence concern, not circularity. Accordingly, the appropriate score is 0.
Assumptions & free parameters
assumptions (2)
- domain assumption The method of moving frames provides a valid formulation for the four time-dependent PDEs on the sphere.
- domain assumption The high-order curvilinear mesh generated by NekMesh achieves negligible geometric approximation error at polynomial order p.
Cite this review
Pith. "Pith review of Numerical study on the effect of geometric approximation error in the numerical solution of PDEs using a high-order curvilinear mesh." pith.science (2026). https://pith.science/paper/VSYM5IBK
@misc{pith2026190809917,
author = {Pith},
title = {Pith review of: Numerical study on the effect of geometric approximation error in the numerical solution of PDEs using a high-order curvilinear mesh},
year = {2026},
howpublished = {\url{https://pith.science/paper/VSYM5IBK}},
note = {Machine review of arXiv:1908.09917}
}
read the original abstract
When time-dependent partial differential equations (PDEs) are solved numerically in a domain with curved boundary or on a curved surface, mesh error and geometric approximation error caused by the inaccurate location of vertices and other interior grid points, respectively, could be the main source of the inaccuracy and instability of the numerical solutions of PDEs. The role of these geometric errors in deteriorating the stability and particularly the conservation properties are largely unknown, which seems to necessitate very fine meshes especially to remove geometric approximation error. This paper aims to investigate the effect of geometric approximation error by using a high-order mesh with negligible geometric approximation error, even for high order polynomial of order p. To achieve this goal, the high-order mesh generator from CAD geometry called NekMesh is adapted for surface mesh generation in comparison to traditional meshes with non-negligible geometric approximation error. Two types of numerical tests are considered. Firstly, the accuracy of differential operators is compared for various p on a curved element of the sphere. Secondly, by applying the method of moving frames, four different time-dependent PDEs on the sphere are numerically solved to investigate the impact of geometric approximation error on the accuracy and conservation properties of high-order numerical schemes for PDEs on the sphere.
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.