Pith. sign in

REVIEW 3 major objections 5 minor 9 references

A quantitative approach to the regularity of a Riemannian surface

T0 review · 3 major / 5 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 A Riemannian surface is intrinsically $\alpha$-regular exactly when it is extrinsically $\alpha$-regular, and the two regularity radii agree up to constants.

desk verdict A genuinely new quantitative equivalence between two surface regularity notions, of real use to geometric analysts; the one direction that leans on the author's separate preprint needs referee checking. read the letter →

arxiv 2507.21921 v1 pith:PHVGBVQY submitted 2025-07-29 math.DG math.MG

Matan Eilat This is my paper
classification math.DGmath.MG MSC 53C2153A05
keywords intrinsicregularityradiusextrinsicGausscurvatureH\"oldernormC^{2\alpha}isothermalcoordinatesRiemanniansurfacesquantitativeharmonic
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 introduces two quantitative notions of what it means for a Riemannian surface to be $C^{2,\alpha}$ at a point, for $0<\alpha<1$: an intrinsic radius controlled by the $\alpha$-H\"older norm of the Gauss curvature on a geodesic disc, and an extrinsic radius controlled by the existence of a local chart in which the metric coefficients are $C^{2,\alpha}$-close to the Euclidean metric. The paper's aim is to prove that these two radii are equivalent, bounding each other up to constants that do not depend on the surface. A sympathetic reader would care because equivalence means a curvature-based and a coordinate-based description of local regularity capture the same geometric information, so results phrased in one language transfer to the other at the cost of a bounded factor.

What carries the argument

For the intrinsic-from-extrinsic direction the paper works in an $\varepsilon$-restricted chart with $\varepsilon=0.01$, writes the Gauss curvature through the standard Christoffel-symbol formula, uses a Laplacian comparison estimate to show the chart domain is strongly convex in the Riemannian sense and convex in the Euclidean sense, and then converts $C^{2,\alpha}$ bounds on the metric coefficients into $C^{0,\alpha}$ bounds on $K$ via the inclusion of $C^1$ functions in $C^{0,\alpha}$. For the extrinsic-from-intrinsic direction the carrying object is an isothermal coordinate chart $z:B\to\delta D$ with conformal factor $\varphi$, whose sup-norm and bi-Lipschitz control by $\delta^2\kappa$ is taken from an earlier preprint. Combining the conformal-factor identity $K=-\Delta(\log\varphi)/(2\varphi)$ with weighted second-order elliptic estimates for this equation yields $\|\varphi-1\|^*_{2,\alpha}\lesssim \delta^2\|K\|_{0,\alpha}$, and a final coordinate rescaling arranges $\varphi(0)=1$, meeting the extrinsic definition on a slightly smaller disc.

What would settle it

Take a rotationally symmetric surface whose Gaussian curvature is a tall narrow bump of height $\kappa$ and width $r\ll\delta$ within a unit disc, compute $\rho_{\mathrm{int}}$ and $\rho_{\mathrm{ext}}$ as functions of $\kappa$, $r$, and $\alpha$, and check whether the ratio $\rho_{\mathrm{ext}}/\rho_{\mathrm{int}}$ stays within constants depending only on $\alpha$; the theorem predicts a uniform bound, so a family where the ratio tends to $0$ or $\infty$ while $\delta^2\|K\|_{0,\alpha;B}\le 1$ would refute it.

Watch

Extended reading notes

Core claim

The central claim is the two-sided comparison $\rho_{\mathrm{int}} \ge C_1 \rho_{\mathrm{ext}}$ for all $0<\alpha\le 1$ with a universal constant $C_1$, and $\rho_{\mathrm{ext}} \ge C_2 \rho_{\mathrm{int}}$ for $0<\alpha<1$ with $C_2$ depending on $\alpha$. In the intrinsic definition, $\rho_{\mathrm{int}}$ is the largest $\delta$ such that the geodesic disc $B(p_0,\delta)$ has injectivity radius at least $2\delta$ at every point and $\delta^2\|K\|_{0,\alpha;B}\le 1$; in the extrinsic definition, $\rho_{\mathrm{ext}}$ is the largest $\delta$ such that $B(p_0,\delta)$ is isometric to a Euclidean coordinate chart with $g_{ij}(p_0)=\delta_{ij}$ and $\|g_{ij}-\delta_{ij}\|_{2,\alpha;U}\le 1$. The equivalence makes the global regularity quantity $\|M\|_\alpha=\rho^{-2}$ independent of which radius is chosen, zero exactly when $M$ is the Euclidean plane, and homogeneous of degree one under scaling of the metric.

Load-bearing premise

The reverse inequality rests on the imported isothermal-coordinate estimate [3], which asserts that a disc with $\delta^2$ times the sup norm of its curvature small has an isothermal chart whose conformal factor and distortion are controlled by that same quantity; if this estimate is false or needs stronger hypotheses, the extrinsic-from-intrinsic direction has no proof.

Editorial extensions

If this is right

  • Estimates in either language transfer to the other: a curvature-H\"older bound on a $\delta$-disc yields metric coefficients $C^{2,\alpha}$-close to Euclidean on a $c(\alpha)\delta$-disc, and conversely a $C^{2,\alpha}$ coordinate representation yields a curvature-H\"older bound on a disc of comparable size.
  • A metric that is $C^{2,\alpha}$-controlled in any chart on a $\delta$-disc is $C^{2,\alpha}$-controlled in an isothermal chart on a disc of radius $C\delta$, a quantitative confirmation that isothermal coordinates are no worse than arbitrary smooth coordinates.
  • The global norm $\|M\|_\alpha=\rho^{-2}$ gives a single number measuring how far a surface is from being uniformly $C^{2,\alpha}$; it vanishes exactly for the Euclidean plane and scales naturally with the metric.
  • In two dimensions, the intrinsic radius offers a curvature-based counterpart to the harmonic radius, so compactness or approximation arguments that use one notion can be rephrased with the other.

Reading between the lines

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

  • Because the reverse inequality is proved only for $\alpha<1$, I expect the optimal constant $C_2(\alpha)$ to deteriorate as $\alpha\to 1^-$, and the rate of that deterioration is not addressed by the paper.
  • A natural test of sharpness is to let curvature oscillate at scale smaller than $\delta$ while keeping $\delta^2\|K\|_{0,\alpha}$ bounded; the theorem predicts the coordinate chart can still be controlled, which is a stronger assertion than the classical regularity theorem alone.
  • If the imported isothermal-chart estimate is replaced by a self-contained proof, the extrinsic direction would gain both independence and clarity about which boundary regularity is really needed for the convexity lemma.
  • The equivalence could serve as a definition of a quantitative $C^{2,\alpha}$ scale in surface fitting: a surface with small $\|M\|_\alpha$ near a point should admit local approximations with error controlled by the radius.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces two quantitative notions of α-regularity for a Riemannian surface: an intrinsic α-regularity radius ρ_int, controlled by the non-dimensional C^{0,α} norm of the Gauss curvature and an injectivity-radius lower bound, and an extrinsic α-regularity radius ρ_ext, controlled by the C^{2,α} closeness of the metric coefficients to the Euclidean metric in some chart. The main results, Theorems 1.3 and 1.4, assert that ρ_int ≥ C1·ρ_ext for 0<α≤1 with a universal constant C1, and ρ_ext ≥ C2·ρ_int for 0<α<1 with a constant C2 depending on α. Section 2 proves the first direction through a reduction to an ε-restricted chart (Proposition 2.2), curvature bounds (Lemma 2.3), Euclidean convexity of the chart domain (Lemma 2.4), and a Hölder estimate for the curvature (Proposition 2.5). Section 3 proves the second direction using an isothermal coordinate chart, a quantitative conformal-factor bound imported from the author's preprint [3] (Theorem 3.1), and weighted Schauder and interpolation estimates (Propositions 3.2–3.5).

Significance. If the theorems are correct, the paper provides a quantitative bridge between curvature regularity and metric-coefficient regularity in two dimensions, with potential applications to surface fitting and approximation theory. The proof of Theorem 1.3 is self-contained and the elementary estimates in Section 2 are carefully executed; the reduction to an ε-restricted chart and the curvature Hölder estimate are clean and appear sound. The proof of Theorem 1.4, however, rests essentially on Theorem 3.1 from the author's separate preprint [3], which is not proved in this manuscript; the extrinsic-to-intrinsic direction is therefore conditional on the correctness and completeness of that companion result. The conceptual framework and the self-contained direction are valuable, but the paper as it stands does not fully deliver the advertised equivalence without an independent derivation of the imported isothermal-chart estimate.

major comments (3)
  1. [§3, Theorem 3.1 and Proposition 3.5] Theorem 1.4 and Proposition 3.5 depend critically on Theorem 3.1, imported from the author's preprint [3]. Specifically, Proposition 3.5 uses the sup-norm bound |log φ| ≤ 8δ²κ, the distance comparability exp(−4δ²κ) ≤ d_g/d_{g0} ≤ exp(4δ²κ), and the existence of the chart z: B → δD with z(p0)=0 and z(∂B)=δS1. None of these statements is proved in the present manuscript; Theorem 3.1 is quoted from [3, Cor. 1.2], and its hypotheses are not verified beyond the brief statement. If [3] has not been independently verified, or if its constants or normalization differ from those needed here, the chain of Schauder estimates in Proposition 3.5 cannot start and the extrinsic-to-intrinsic direction fails. The manuscript should either include a proof of Theorem 3.1 in an appendix or explicitly state that the main theorem is conditional on an unpublished companion paper; it should also verify that the hypotheses of [3, Cor. 1.2] are satisfied, namely that δ²κ < π²/8, which is met here because δ²∥K∥_{0,α;B} ≤ 0.01.
  2. [§3, proof of Theorem 1.4] The proof assumes the existence of an isothermal coordinate chart z0 : B0 → δ0D with z0(p0)=0 and z0(∂B0)=δ0S1, and Proposition 3.5 makes the same assumption. However, Theorem 3.1 as stated is conditional: it begins 'Let z : B → δD be an isothermal coordinate chart such that...' and does not assert that such a chart exists. Existence of a conformal equivalence from a geodesic disk to a Euclidean disk follows from the uniformization theorem and smoothness of the boundary, but the paper neither states nor cites this. This is a load-bearing gap in the proof of Theorem 1.4; it should be repaired by adding an explicit existence statement with a proof or reference.
  3. [§2, Lemma 2.4] The proof of Lemma 2.4 assumes that the boundary ∂U is regular enough to be parametrized as a unit-speed curve γ : (−τ, τ) → ∂U and to support the Euclidean acceleration computation. This regularity follows from the fact that ∂B is a smooth distance sphere under the injectivity-radius hypothesis, but the manuscript does not mention this. The issue is easily fixable, but as written the proof begins with an unjustified assumption about the boundary.
minor comments (5)
  1. [§3, Theorem 3.1] The wording 'Let z : B → δD be an isothermal coordinate chart...' is ambiguous; if the theorem is intended to assert existence, it should say 'There exists an isothermal coordinate chart z...'. This is related to Major Comment 2.
  2. [§2, Lemma 2.4] Consider stating explicitly that ∂B is a smooth curve because the injectivity radius at p0 is at least 2δ, so the exponential map is a diffeomorphism on the tangent ball of radius 2δ.
  3. [§3, proof of Theorem 1.4] The quantity δ0 is written as 'δ/10√C0'; this should be typeset as δ/(10√C0) to avoid ambiguity.
  4. [§3, proof of Theorem 1.4] The conversion from the weighted norm on δ0D to the unweighted norm on U is stated with constant 8, but summing the factors for the C^0, C^1, C^2 and C^{2,α} pieces gives a larger constant (up to about 15 for α=1). The argument still works because the final bound only needs to be <1, but the displayed constant should be corrected or replaced by an unspecified absolute constant.
  5. [References] The name 'Peterson' in reference [6] should be 'Petersen'.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction found: the intrinsic and extrinsic regularity radii are defined from different data, and both implications are genuine estimates; the main caveat is that the intrinsic-to-extrinsic direction (Theorem 1.4) relies on a load-bearing bound imported from the author's own preprint [3], which is a verification risk rather than a by-construction circularity.

full rationale

The two radii quantify different data: Definition 1.1 (intrinsic) is stated in terms of δ²‖K‖_{0,α;B} ≤ 1 with an injectivity-radius condition, while Definition 1.2 (extrinsic) is stated in terms of a chart with g_ij(x₀)=δ_ij and ‖g_ij−δ_ij‖_{2,α;U} ≤ 1. Neither definition references the other, and I found no step in either direction where the conclusion is fed back as an input. Theorem 1.3 (extrinsic ⇒ intrinsic) is self-contained: Proposition 2.2 zooms to ε-restricted coefficients, Lemma 2.3 bounds sup|K| by a direct Christoffel-symbol computation, Lemma 2.4 gives Euclidean convexity of U (with an unstated C²-boundary assumption on ∂U, a fixable gap), and Proposition 2.5 bounds the Hölder norm of K from the C^{2,α} norms of g_ij. That is a direct a-priori estimate, not a repackaging of the definition. Theorem 1.4 (intrinsic ⇒ extrinsic) is the delicate direction: its proof uses Proposition 3.5, which starts from Theorem 3.1, imported verbatim from the author's own preprint: 'Theorem 3.1. [3, Corollary 1.2]... sup_B |log φ| ≤ 8δ²κ, and exp(−4δ²κ) ≤ d_g(p,q)/d_g0(z(p),z(q)) ≤ exp(4δ²κ).' Proposition 3.5 then states: 'According to Theorem 3.1 ... ∥φ−1∥₀ ≤ 2∥log φ∥₀ ≤ 16·δ²∥K∥₀', and uses the bi-Lipschitz comparability to identify Hölder norms of K on B and δD. This is genuinely load-bearing self-citation: if [3, Cor. 1.2] is false or needs hypotheses beyond those stated, the Schauder chain cannot start and Theorem 1.4 collapses. Nevertheless, this is not a circular reduction: the cited result (C⁰ control of the conformal factor plus a bi-Lipschitz estimate) is strictly weaker than Theorem 1.4's conclusion (C^{2,α} control of φ−1), and the gap is filled by Poisson/Schauder estimates from Gilbarg–Trudinger [4] applied to Liouville's equation K = −Δ(log φ)/(2φ). The stated hypotheses of [3, Cor. 1.2] (injectivity radius ≥ 2δ, |K| ≤ κ, δ²κ < π²/8) do not include the target result, and no fitted parameter is renamed as a prediction anywhere in the paper. Accordingly the import qualifies as real evidence under the stated rules, with the residual risk being an unverified preprint — a correctness and verification concern, not an equivalence-by-construction. Score 2 reflects this: a significant but non-circular load-bearing self-citation.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard elliptic theory (Schauder estimates, interpolation) and comparison geometry, plus a single specialized input: Theorem 3.1 from the author's earlier preprint [3]. The auxiliary constant ε=0.01 is chosen by hand but only affects constants. No new entities are postulated.

free parameters (1)
  • epsilon = 0.01
    Chosen by hand as a sufficiently small norm bound to make the curvature estimates in Section 2 work; a universal constant, not fitted to data.
assumptions (5)
  • standard math Schauder estimates and interpolation inequalities for Poisson's equation (Gilbarg-Trudinger, used in Proposition 3.5)
    Imported from [4]; standard PDE theory.
  • standard math Laplace comparison for the distance function on a surface with curvature bounded above and below
    Used in Lemma 2.4 to bound Δr and prove Euclidean convexity of the chart domain.
  • standard math Whitehead's theorem on strong convexity of small geodesic balls
    Used in Lemma 2.3 to conclude the Riemannian disc is strongly convex.
  • domain assumption Theorem 3.1 from [3]: existence of an isothermal chart z: B → δD with z(p0)=0, z(∂B)=δS1 and quantitative bounds on the conformal factor and the bi-Lipschitz constant in terms of δ²κ
    This is the key unproved input in the present paper; it is the author's own preprint and is essential for Proposition 3.5 and Theorem 1.4.
  • ad hoc to paper The boundary of the chart domain U in Lemma 2.4 is regular enough to be parametrized as a unit-speed curve and to support the Euclidean acceleration computation
    No justification is given for the C^2 regularity of ∂U; the proof relies on it implicitly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A quantitative approach to the regularity of a Riemannian surface." pith.science (2026). https://pith.science/paper/PHVGBVQY

@misc{pith2026250721921,
  author       = {Pith},
  title        = {Pith review of: A quantitative approach to the regularity of a Riemannian surface},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PHVGBVQY}},
  note         = {Machine review of arXiv:2507.21921}
}
abstract

We introduce two definitions with the purpose of quantifying the concept of a $C^{2,\alpha}$ surface for $0 < \alpha < 1$. The intrinsic definition is given in terms of the $\alpha$-H\"{o}lder norm of the Gauss curvature function. The extrinsic one relies on the existence of a smooth local representation of the Riemannian metric. We show that these definitions are equivalent up to a constant depending on $\alpha$.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

9 extracted references · 9 canonical work pages

  1. [3]

    The bi-Lipschitz constant of an isothermal coordinate chart

    Eilat, M., The bi-Lipschitz constant of an isothermal coordinate chart. Preprint, arXiv:2505.23591

    work page Pith review arXiv

  2. [1]

    T., Convergence and rigidity of manifolds under Ricci curvature bounds

    Anderson, M. T., Convergence and rigidity of manifolds under Ricci curvature bounds. Invent. Math., vol. 102, no. 2, (1990), 429–445

    work page 1990

  3. [2]

    M., Kazdan, J

    DeTurck, D. M., Kazdan, J. L., Some regularity theorems in Riemannian geometry. Ann. Sci. ´Ecole Norm. Sup. (4), vol. 14, no. 3, (1981), 249–260. 12

    work page 1981

  4. [4]

    S., Elliptic partial differential equations of second order

    Gilbarg, D., Trudinger, N. S., Elliptic partial differential equations of second order. Grundlehren der Mathematischen Wissenschaften, vol. 224, Springer, Berlin-New York, 1977

    work page 1977

  5. [5]

    Hebey, E., Herzlich, M., Harmonic coordinates, harmonic radius and convergence of Rie- mannian manifolds. Rend. Mat. Appl. (7), vol. 17, no. 4, (1997), 569–605

    work page 1997

  6. [6]

    Springer New York (2006)

    Peterson, P., Riemannian Geometry. Springer New York (2006)

    work page 2006

  7. [7]

    J., Lectures on Classical Differential Geometry

    Struik, D. J., Lectures on Classical Differential Geometry. 2nd. ed. Dover Publications (1961)

    work page 1961

  8. [8]

    Whitehead, J. H. C., Convex regions in the geometry of paths.Quart. J. Math., vol. 3, no. 1, (1932), 33–42

    work page 1932

Show all 9 references
  1. [9]

    S., Zhu, M., Bounds on harmonic radius and limits of manifolds with bounded Bakry- ´Emery Ricci curvature

    Zhang, Q. S., Zhu, M., Bounds on harmonic radius and limits of manifolds with bounded Bakry- ´Emery Ricci curvature. J. Geom. Anal., vol. 29, no. 3, (2019), 2082–2123. Department of Mathematics, Weizmann Institute of Science, Rehovot 76100, Israel. e-mail: matan.eilat@weizmann...

    2019

Pith tools

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