Pith. sign in

REVIEW 4 major objections 5 minor 20 references

Domains without dense Steklov nodal sets

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Steklov eigenfunctions on a dense family of analytic domains can remain nonzero on a fixed-radius ball, so their nodal sets are not dense at shrinking scales.

desk verdict A genuinely new negative answer to a named Steklov open problem, with a coherent proof built on one load-bearing imported estimate that deserves referee scrutiny. read the letter →

arxiv 1908.03307 v1 pith:PDUQFWJM submitted 2019-08-09 math.AP math.SP

Oscar Bruno , Jeffrey Galkowski This is my paper · ORCID
classification math.APmath.SP MSC 35P1535J0535B0530C35
keywords SteklovproblemnodalsetstunnelingconditionBBLCNdomainsconformalmappingboundaryFouriercoefficientshigh-frequencyasymptoticsspectralgeometry
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

This paper answers an open question about Steklov eigenfunctions: whether their zero sets must become dense near the high-frequency limit. It shows the answer is no. For any bounded simply-connected planar domain with analytic boundary, one can perturb the boundary by an arbitrarily small analytic deformation so that, in the new domain, every Steklov eigenfunction is nonzero on some ball of a fixed radius, with the location of the ball allowed to depend on the eigenvalue. Therefore, for these domains, the nodal set is not dense at scale $\sigma^{-1}$, and in a fixed interior region the eigenfunctions oscillate no faster than a bounded frequency. The proof works by showing that a dense class of domains, those with boundary-band-limited conformal maps, satisfy a tunneling condition forcing the interior eigenfunction to be dominated by finitely many Fourier modes.

What carries the argument

The central device is the tunneling condition, defined via boundary Fourier coefficients of the pulled-back eigenfunction $u=\varphi\circ f$ on the unit disk: for any $K$ there is $C_0$ such that $|\hat u(k)|\le C_0^{|k-m|}A_m$ for $|k|\le K\sigma$, where $A_m=(\sum_{k=m-m_0}^{m+m_0}|\hat u(k)|^2)^{1/2}$. Lemma 4.1 turns this into a lower bound $e^{-C\sigma}\|\hat u\|_{\ell^2}\le A_m$ on the low-frequency mass, so the interior harmonic extension cannot become exponentially negligible. Theorem 3 proves the tunneling condition for BBLCN domains, i.e. domains whose conformal mapping $f$ has $|\partial_z f|$ boundary-band-limited and nonconstant on $\partial\mathbb{D}$, for example $f(z)=\int p(w)^2\,dw$ with polynomial $p$ having no roots in the disk. Corollary 2.2 approximates any analytic simply-connected domain arbitrarily closely in $C^k$ by such domains. Theorem 2 then uses the $r^{|k|}$ decay of Fourier harmonics to dominate high modes, leaving a low-frequency polynomial that cannot vanish throughout any fixed-radius ball.

What would settle it

Compute boundary Fourier coefficients of high Steklov eigenfunctions on a non-BBLCN analytic domain, such as the ellipse with semiaxes 2 and 1; if for some sequence $\sigma_j\to\infty$ the mass in the window $|k-m|\le 2\sigma_j$ falls below $(1-Ce^{-\sigma_j/C})$ of the total mass, the exponential-concentration input to Lemma 4.1 fails and the proof chain collapses. A direct check of the paper's conclusion would be to search for a high eigenfunction whose zero set intersects every ball of radius $r_1$ inside the domain for every $r_1>0$.

Watch

Extended reading notes

Core claim

The paper's central discovery is that high-frequency Steklov eigenfunctions can be frozen inside a domain. Theorem 1 states that, starting from any bounded simply-connected analytic domain $\Omega_0$, one can make an arbitrarily small analytic boundary perturbation (of size $\varepsilon$ in $C^k$) to obtain a domain $\Omega_1$ with a point $x_0$ and radii $0<r_1<r_0$ such that every Steklov eigenfunction of every eigenvalue $\sigma$ is nonzero on some ball $B(x_\sigma,r_1)$ inside $B(x_0,r_0)$. The point $x_\sigma$ may depend on $\sigma$, but $r_1$ does not. Since the expected oscillation scale is $\sigma^{-1}$, this is incompatible with the nodal set being dense at that scale, giving a negative answer to Open Problem 10(i) of [GP17]. The refined Theorem 2 shows that on a small ball the eigenfunction is approximated in $C^N$ by finitely many Fourier modes, with relative error $C_N(\delta m^{-N-m_0-1}+e^{-c\sigma})$, and Theorem 3 identifies a dense class of domains, the BBLCN domains, for which the necessary tunneling estimate holds.

Load-bearing premise

The whole chain depends on an imported exponential concentration estimate for the boundary Fourier coefficients of a Steklov eigenfunction; if that estimate fails for some high eigenvalue, the low-frequency mass bound and the resulting constant-sign ball are not established.

Editorial extensions

If this is right

  • Open Problem 10(i) of [GP17] has a negative answer: in the constructed domains, Steklov nodal sets are not dense at scale $\sigma^{-1}$, and a fixed-radius ball of constant sign exists for every eigenfunction.
  • The class of simply-connected analytic domains exhibiting this behavior is dense in the $C^k$ boundary topology, so the phenomenon is not confined to special shapes.
  • On such domains, interior Steklov eigenfunctions are well approximated by finitely many Fourier modes: the $C^N$ error on a small ball is bounded by $C_N(\delta m^{-N-m_0-1}+e^{-c\sigma})$ relative to the $L^2$ norm.
  • The eigenfunctions have bounded frequency on a fixed interior neighborhood, so their oscillation rate does not grow with $\sigma$ in that region.
  • Every analytic simply-connected domain can be approximated arbitrarily closely by a BBLCN domain, and every BBLCN domain is tunneling.

Reading between the lines

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

  • If the tunneling conjecture stated in the paper is true, the fixed-radius sign-ball behavior would extend to every analytic non-circular domain, making non-density generic; the ellipse and kite numerics already show the expected opening.
  • The proof reframes interior Steklov decay as an exponential tunneling effect, suggesting that interior Steklov eigenfunctions localize away from some regions in a way high-energy Laplace eigenfunctions do not; a comparison of nodal length asymptotics on the same domain could test this distinction.
  • A parameter-free extension would be to run the BBLCN construction with rational maps instead of polynomial squares, allowing the method to reach multiply connected or higher-genus geometries once the band-limited algebra is verified.
  • A quick numerical check of the mechanism is to perturb the unit disk by one low-order Fourier mode and track the largest sign-definite ball as $\sigma$ grows; the theory predicts the radius stays bounded below, whereas for the exact disk it shrinks like $\sigma^{-1}$.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper studies Steklov eigenfunctions in bounded simply-connected planar domains with analytic boundaries. It constructs a dense family of such domains for which, for every sufficiently large Steklov eigenvalue, every eigenfunction has a ball of sigma-independent radius on which it does not vanish, thereby giving a negative answer to Open Problem 10(i) of Girouard and Polterovich (J. Spectr. Theory 2017). The proof combines an approximation theorem by boundary-band-limited conformal domains (Corollary 2.2), a new 'tunneling' condition (Definition 1.1) shown to hold for BBLCN domains (Theorem 3), an exponential lower bound on low-frequency Fourier mass (Lemma 4.1), and a finite-mode approximation of interior eigenfunctions (Theorem 2). The paper also contains numerical experiments for elliptical and kite-shaped domains.

Significance. If the main theorem holds, the result is significant: it establishes that Steklov eigenfunctions on a dense class of analytic planar domains can have slowly oscillating, sign-definite regions of fixed size, in stark contrast to high-energy Laplace eigenfunctions. The tunneling mechanism is a new and interesting idea, and the construction of BBLCN approximating domains is explicit and uses no fitted parameters. The proof of Theorem 3 is a clean recurrence argument. However, the central chain depends on an imported spectral concentration estimate from [GT19] and on a quantifier in the statement of Theorem 1 that is stronger than what the proof appears to establish; both points need to be addressed before the paper can be accepted.

major comments (4)
  1. [Section 1, Theorem 1] The statement 'there exists a point x_sigma in B(x0,r0) such that ... each Steklov eigenfunction phi_sigma of eigenvalue sigma satisfies |phi_sigma|>0 on B(x_sigma,r1)' requires a single x_sigma common to all eigenfunctions in the eigenspace of sigma. The proof, around equations (4.4)-(4.10), chooses x0 from the maximum of the finite-mode approximation tilde u_{sigma,delta,m} for one eigenfunction, so it proves the conclusion only for that eigenfunction. If the Steklov spectrum of the constructed domain has an eigenvalue of multiplicity at least two, no common point can work: given two independent eigenfunctions positive at a proposed x_sigma, a suitable linear combination vanishes at x_sigma. Thus the theorem as stated is either false for eigenspaces of dimension greater than one or at least not proved. Please restate the theorem in the per-eigenfunction form used in the abstract, or add a genericity argument ensuring the constructed domains have simple Steklov spectrum.
  2. [Section 4, Lemma 4.1] The proof of Lemma 4.1 imports the estimate [GT19, Corollary 1.3] in the form sum_{|k-m|<=2sigma}|hat u(k)|^2 >= ||hat u||_{ell^2}^2(1 - C e^{-sigma/C}) without stating its precise hypotheses or proof. This estimate is load-bearing: it is the only source of the exponential lower bound e^{-C sigma}||hat u|| <= A_m, and without it the high-frequency tail estimate in Theorem 2 and the fixed-radius positivity in Theorem 1 fail. Since the corollary is external and is co-authored by the second author, a reader cannot certify it from the text alone. Please state the exact statement, confirm that it applies to the mapped Steklov problem on the unit disk with the weight |partial_z f|, and verify the m=0 case, which is used in the proof of Theorem 2 even though Lemma 4.1 as stated is for m>0.
  3. [Section 4, proof of Lemma 4.1] The displayed application of Lemma 3.1 appears to have an indexing error: the sums are written with factors C_0^{2k} and C_0^{2|k|}, whereas Lemma 3.1 gives the bound C_0^{2|k-m|} for the coefficients in the window |k-m|<=2sigma. With the printed exponents, the subsequent bound by (2(2C_0^{4sigma+2}-1)/(C_0^2-1)) A_m^2 does not follow, particularly when m is large. This is likely a typographical error, but it must be corrected because the lower bound for A_m depends on this step.
  4. [Section 4, proof of Theorem 2 and Section 1, Theorem 1] The proof of Theorem 1 uses the C^1 bound of tilde u_{sigma,delta,m} on a ball B(x0,r_{m,delta}) but the preceding estimates were obtained on B(0,delta). To ensure B(x0,r_{m,delta}) is contained in the unit disk and that the image f(B(x0,r_{m,delta})) lies in Omega_1 with the stated ball inclusion B(x_sigma,r1) subset B(x0,r0), the proof should specify the choice of delta and r_{m,delta} more carefully, e.g. taking 2delta<1 and r_{m,delta}<delta. The current text says r_{m,delta}<delta but does not state the required containment in the domain of definition of the approximation.
minor comments (5)
  1. [Definition 1.1] The tunneling condition does not explicitly quantify over eigenfunctions; please state that the inequality is required for every Steklov eigenfunction u_sigma (or for every member of an orthonormal basis).
  2. [Notation] The notation switches between u_{sigma_j}, u_sigma, and tilde u_{sigma,delta}; please define these consistently near (1.3), (1.7), and Theorem 2.
  3. [Lemma 2.1] The proof approximates w by a polynomial p_epsilon and then factors p_epsilon = beta_0 prod (z-beta_i)^{N_i}. To ensure the zeros satisfy |beta_i|>1 rather than merely |beta_i|>=1, one should note that a small additional perturbation can push boundary zeros slightly outside the closed disk; as written this point is implicit.
  4. [Corollary 2.2] The implicit-function-theorem argument for expressing partial Omega_{epsilon} as a normal graph is terse; for instance, the periodicity and C^k regularity of s(theta) and omega(theta) are asserted rather than shown. A few clarifying sentences would help.
  5. [Figures and tables] There are small presentation errors: Figure 2's caption contains the typo 'eignfunction', and Table 2's caption says 'Same as Figure (1)' where 'Table 1' is meant.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the construction of BBLCN approximants and the tunneling condition are genuine new inputs, and the only self-citation [GT19, Cor 1.3] is an external parameter-free estimate used as a lemma.

full rationale

The paper's derivation chain is: Corollary 2.2 approximates any analytic domain by BBLCN domains using polynomial approximation of ∂_z f; Theorem 3 proves that BBLCN domains satisfy the tunneling condition (Definition 1.1) via the band-limited recurrence (3.2)-(3.3); Lemma 4.1 combines Lemma 3.1 with the imported concentration estimate [GT19, Cor. 1.3] to obtain low-frequency Fourier mass; Theorem 2 derives the finite-mode truncation estimate (1.8) from the tunneling condition; Theorem 1 then combines Theorem 2 with the density of BBLCN domains. Each step is a proof rather than a definitional identification. The tunneling condition is a genuine hypothesis about exponential concentration of boundary Fourier coefficients near a frequency window; it is not defined in terms of nodal-set non-density, and the conclusion of positivity on a fixed-size ball is not inserted into the condition. The only self-citation is [GT19, Cor. 1.3], a published parameter-free estimate on Steklov eigenfunctions; the paper uses it as an external lemma, it is not derived from the target result, and it does not presuppose the existence of domains with non-dense nodal sets. Consequently, the citation is load-bearing for the proof but not circular. Internal wording issues, such as Lemma 4.1 being stated for m>0 but later applied with m=0, are correctness or typographical concerns rather than circularity. The central claim does not reduce to a fit, a renamed known result, or a self-citation chain.

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

The central proof depends on standard complex analysis, the implicit function theorem, and one imported spectral estimate [GT19, Corollary 1.3]. No free parameters are fitted. The tunneling condition is a new definition proved for BBLCN domains, not an assumed hypothesis in the final theorem.

assumptions (3)
  • standard math Riemann mapping theorem and Bell-Krantz [BK87] smooth extension of conformal maps to the boundary for analytic simply-connected domains.
    Used in Section 1 to define the conformal mapping f: D -> Ω and in Corollary 2.2 to justify approximation by BBLCN domains.
  • standard math Implicit function theorem used to show the perturbed curve ∂Ω_ε is a normal graph over ∂Ω (Corollary 2.2).
    Entered in the proof of Corollary 2.2 via the function F(t,θ,ω,s); the theorem guarantees unique s(θ), ω(θ).
  • domain assumption [GT19, Corollary 1.3]: exponential concentration of boundary Fourier coefficients of Steklov eigenfunctions in the window |k-m|≤2σ.
    Imported in Lemma 4.1; it is the key external input that turns the tunneling decay into an exponential lower bound on low-frequency mass.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Domains without dense Steklov nodal sets." pith.science (2026). https://pith.science/paper/PDUQFWJM

@misc{pith2026190803307,
  author       = {Pith},
  title        = {Pith review of: Domains without dense Steklov nodal sets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PDUQFWJM}},
  note         = {Machine review of arXiv:1908.03307}
}
abstract

This article concerns the asymptotic geometric character of the nodal set of the eigenfunctions of the Steklov eigenvalue problem $$ -\Delta \phi_{\sigma_j}=0,\quad\text{ on }\Omega,\qquad\qquad \partial_\nu \phi_{\sigma_j}=\sigma_j \phi_{\sigma_j}\quad \text{ on }\partial\Omega $$ in two-dimensional domains $\Omega$. In particular, this paper presents a dense family $\mathcal{A}$ of simply-connected two-dimensional domains with analytic boundaries such that, for each $\Omega\in \mathcal{A}$, the nodal set of the eigenfunction $\phi_{\sigma_j}$ "is $not$ dense at scale $\sigma_j^{-1}$". This result addresses a question put forth under "Open Problem 10" in Girouard and Polterovich, J. Spectr. Theory, 321-359 (2017). In fact, the results in the present paper establish that, for domains $\Omega\in \mathcal{A}$, the nodal sets of the eigenfunctions $\phi_{\sigma_j}$ associated with the eigenvalue $\sigma_j$ have starkly different character than anticipated: they are not dense at any shrinking scale. More precisely, for each $\Omega\in \mathcal{A}$ there is a value $r_1>0$ such that for each $j$ there is $x_j\in \Omega$ such that $\phi_{\sigma_j}$ does not vanish on the ball of radius $r_1$ around $x_j$.

Figures

Figures reproduced from arXiv: 1908.03307 by the authors.

Figure 1
Figure 1. Fixed-sign sets for Steklov eigenfunctions over the elliptical domain Ω = x 2 + y 2 1.012 = 1. The yellow and blue regions indicate the subsets over which the eigenfunctions are positive and negative, respectively. The left and right images correspond to the eigenvalues σ20 = 9.9502 and σ30 = 14.9253, respectively. For a circle the nodal lines coincide with a set of j uniformly arranged radial lines from the center … view at source ↗
Figure 2
Figure 2. Steklov eigenfunctions on the domain Ω whose mapping function, which is given by equation (2.3), maps the center of the disk to the point z0 = (0.8, 0) (marked by red asterisks in the figures). The corresponding Steklov eigenvalues are given by σ16 = 7.9642 (top left), σ40 = 19.8173 (top right), and σ60 = 29.8197 (bottom left). Note that, according to Corollary 2.2 the set Ω is a BBLCN approx￾imation to the disk. As… view at source ↗
Figure 3
Figure 3. The function λ for an ellipse (left) and a kite-shaped domain (right). 5. Numerical Formulation 5.1. Integral representation. Let Ω ⊂ R 2 denote a domain with, say, a C 2 boundary, and let S[φ](x) := Z ∂Ω G(x, y)φ(y)dS(y), x ∈ R 2 , G(x, y) = − 1 2π log |x − y|, denote the Single Layer Potential (SLP) for a given density φ : ∂Ω → R in a certain Banach space H of functions. Both Sobolev and continuous spaces H of fun… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Density-plots (first and third rows) and fixed-sign sets (second and forth rows) for Steklov eigenfunctions over the elliptical domain (6.1). The eigenfunctions of orders 57 and 81 demonstrate the onset of the asymptotic character. In particular, regions of asymptotica…
Figure 5
Figure 5. Figure 5: Density-plots (first row) and fixed-sign sets (second rows) for Steklov eigenfunctions over the kite-shaped domain (6.2). 5.3. Exponential decay and verification of Cauchy’s theorem. Tables 1 and 2 demonstrate the validity of equation (5.8) (since in both cases the res…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 19 canonical work pages

  1. [1]

    Eldar Akhmetgaliyev. Fast numerical methods for mixed, singular Helmholtz boundary value problems and Laplace eigenvalue problems-with applications to antenna design, sloshing, electromagnetic scattering and spectral geometry . PhD thesis, California Institute of Technology, 2016

    work page 2016

  2. [2]

    Bell and Steven G

    Steven R. Bell and Steven G. Krantz. Smoothness to the boundary of conformal maps. Rocky Mountain J. Math. , 17(1):23--40, 1987

    work page 1987

  3. [3]

    Nodal sets of S teklov eigenfunctions

    Katar\' i na Bellov\' a and Fang-Hua Lin. Nodal sets of S teklov eigenfunctions. Calc. Var. Partial Differential Equations , 54(2):2239--2268, 2015

    work page 2015

  4. [4]

    Bender and Steven A

    Carl M. Bender and Steven A. Orszag. Advanced mathematical methods for scientists and engineers. I . Springer-Verlag, New York, 1999. Asymptotic methods and perturbation theory, Reprint of the 1978 original

    work page 1999

  5. [5]

    Spectral geometry of the S teklov problem (survey article)

    Alexandre Girouard and Iosif Polterovich. Spectral geometry of the S teklov problem (survey article). J. Spectr. Theory , 7(2):321--359, 2017

    work page 2017

  6. [6]

    Alexandre Girouard, Leonid Parnovski, Iosif Polterovich, and David A. Sher. The S teklov spectrum of surfaces: asymptotics and invariants. Math. Proc. Cambridge Philos. Soc. , 157(3):379--389, 2014

    work page 2014

  7. [7]

    Polynomial upper bound on interior Steklov nodal sets

    Bogdan Georgiev and Guillaume Roy-Fortin. Polynomial upper bound on interior S teklov nodal sets. arXiv:1704.04484 , 2017

    work page Pith review arXiv 2017

  8. [8]

    Jeffrey Galkowski and John A. Toth. Pointwise bounds for S teklov eigenfunctions. J. Geom. Anal. , 29(1):142--193, 2019

    work page 2019

Show all 20 references
  1. [9]

    Hislop and Carl V

    Peter D. Hislop and Carl V. Lutzer. Spectral asymptotics of the D irichlet-to- N eumann map on multiply connected domains in R^d . Inverse Problems , 17(6):1717--1741, 2001

    2001

  2. [10]

    Linear integral equations , volume 82 of Applied Mathematical Sciences

    Rainer Kress. Linear integral equations , volume 82 of Applied Mathematical Sciences . Springer, 3 edition, 2014

    2014

  3. [11]

    Sloshing, steklov and corners: Asymptotics of sloshing eigenvalues

    Michael Levitin, Leonid Parnovski, Iosif Polterovich, and David A Sher. Sloshing, steklov and corners: Asymptotics of sloshing eigenvalues. arXiv preprint arXiv:1709.01891 , 2017

    2017 arXiv

  4. [12]

    Strongly elliptic systems and boundary integral equations

    William McLean and William Charles Hector McLean. Strongly elliptic systems and boundary integral equations . Cambridge university press, 2000

    2000

  5. [13]

    Nodal length of S teklov eigenfunctions on real-analytic R iemannian surfaces

    Iosif Polterovich, David Sher, and John Toth. Nodal length of S teklov eigenfunctions on real-analytic R iemannian surfaces. to appear in J. Reine Angew. Math

  6. [14]

    S. E. Shamma. Asymptotic behavior of S tekloff eigenvalues and eigenfunctions. SIAM J. Appl. Math. , 20:482--490, 1971

    1971

  7. [15]

    Sogge, Xing Wang, and Jiuyi Zhu

    Christopher D. Sogge, Xing Wang, and Jiuyi Zhu. Lower bounds for interior nodal sets of S teklov eigenfunctions. Proc. Amer. Math. Soc. , 144(11):4715--4722, 2016

    2016

  8. [16]

    Michael E. Taylor. Partial differential equations II . Q ualitative studies of linear equations , volume 116 of Applied Mathematical Sciences . Springer, New York, second edition, 2011

    2011

  9. [17]

    A lower bound for the nodal sets of S teklov eigenfunctions

    Xing Wang and Jiuyi Zhu. A lower bound for the nodal sets of S teklov eigenfunctions. Math. Res. Lett. , 22(4):1243--1253, 2015

    2015

  10. [18]

    Hausdorff measure of nodal sets of analytic S teklov eigenfunctions

    Steve Zelditch. Hausdorff measure of nodal sets of analytic S teklov eigenfunctions. Math. Res. Lett. , 22(6):1821--1842, 2015

    2015

  11. [19]

    Geometry and interior nodal sets of S teklov eigenfunctions

    Jiuyi Zhu. Geometry and interior nodal sets of S teklov eigenfunctions. arXiv:1510.07300 , 2015

    2015 arXiv

  12. [20]

    Interior nodal sets of S teklov eigenfunctions on surfaces

    Jiuyi Zhu. Interior nodal sets of S teklov eigenfunctions on surfaces. Anal. PDE , 9(4):859--880, 2016

    2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.