The Hausdorff measure of the boundary of the Brownian disk

Alexis Metz--Donnadieu

REVIEW 2 major objections 6 minor 17 references

The Hausdorff measure of the boundary of the Brownian disk

T0 review · 2 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The uniform measure on the boundary of the Brownian disk equals, up to a deterministic constant, the Hausdorff measure with gauge s^2 log log(1/s).

desk verdict Settles the boundary version of Le Gall's Hausdorff measure theorem with genuinely new gauge and Bessel estimates; one fixable gap in the final constancy argument. read the letter →

arxiv 2411.16541 v2 pith:CAYQ2VF5 submitted 2024-11-25 math.PR

Alexis Metz--Donnadieu This is my paper · ORCID

classification math.PR MSC 28A7860D0560G57

keywords Browniandiskhalf-planeHausdorffmeasureuniformgaugefunctionBesselprocessrandommetricspacererootinginvariance

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 proves that the natural uniform measure on the boundary of the Brownian disk is not an extra piece of data: up to a deterministic constant, it is the Hausdorff measure built from the gauge function h(s)=$s^{2}$ log log(1/s) and the metric that the Brownian disk induces on its boundary. The same statement holds for the boundary of the Brownian half-plane. If correct, the intrinsic metric of these random surfaces determines their boundary measure, so one can recover the boundary measure from distances alone. This answers, for the boundary, the question previously resolved for the volume measure of the Brownian sphere.

What carries the argument

The proof combines a density-comparison criterion for Hausdorff measures with the rerooting invariance of the Brownian disk. First, Bessel-process occupation-time estimates (with the gauge h(s)=$s^{2}$ log log(1/s)) show that the Hausdorff measure m_h^D and Lebesgue measure on the boundary are mutually absolutely continuous, with Radon–Nikodym density bounded between two constants. Then rerooting invariance—the law of the disk viewed from a uniform boundary point is the same for every boundary point—transfers the value ψ'(0)=limsup_{ε↓0} m_h^D([0,ε])/ε to every point, forcing the density to be the deterministic constant κ. A zero–one law for the germ algebra in the Brownian half-plane makes κ deterministic and identifies it with the analogous half-plane limsup.

What would settle it

For a single realization, fix two boundary points x and y and compare limsup_{ε↓0} m_h^D(B_ε(x))/h(ε) with the corresponding limsup at y; the theorem predicts both equal the same constant κ, so observing different finite limsup values at two points with positive probability would disprove the claim.

Watch

Extended reading notes

Core claim

Theorem 0.1 asserts that there exists a deterministic constant κ in (0, ∞) such that almost surely the uniform measure on the boundary ∂D of the free Brownian disk coincides with κ m_h, where m_h is the Hausdorff measure associated with the metric restricted to the boundary and the gauge function h(s)=$s^{2}$ log log(1/s). Because the boundary is naturally identified with the circle, the uniform measure is Lebesgue measure, and the theorem shows that it is completely determined by the metric. Corollary 0.2 extends the same statement to the boundary of the Brownian half-plane with the same constant κ, using the known coupling between the two objects.

Load-bearing premise

The proof relies on the assertion that the Radon–Nikodym density of m_h^D with respect to Lebesgue measure is almost surely constant, obtained by identifying the pointwise derivative at the root ψ'(0) with the deterministic limsup κ and then using rerooting invariance to transfer that value to every boundary point; this step is only sketched.

Editorial extensions

If this is right

The boundary measure of the Brownian disk and half-plane is intrinsic: any isometry of these metric spaces preserves the uniform boundary measure.
The same gauge function h(s)=s^2 log log(1/s) and the same constant κ apply to both the disk and the half-plane, so the local boundary geometry has a single universal scaling law.
Together with the analogous result for the Brownian sphere volume, all natural uniform measures in Brownian geometry—both volume and boundary—are Hausdorff measures determined by the metric.
The method gives a template for deriving exact gauge functions of Hausdorff measures on other Brownian or stable surfaces with boundary.

Reading between the lines

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

The exact value of κ is left unidentified; a numerical estimate could in principle be extracted from the explicit constants in the Bessel occupation-time limsup of Ciesielski–Taylor.
A stronger pointwise statement likely holds—that the limsup of m_h^D(B_r(x))/h(r) equals κ at every boundary point, not just almost everywhere—but the paper's argument only establishes the density is constant almost everywhere.
A direct proof that ψ(s)=m_h^D([0,s]) is differentiable at 0 would eliminate the sketched step in Theorem 3.4 and make the rerooting transfer fully explicit.
The same gauge function might also describe Hausdorff measures on boundaries of other random planar maps with boundary, such as stable maps, where a version of this argument could be adapted.

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

Here's what you should know about this paper. It proves that the uniform measure on the boundary of the Brownian disk is a Hausdorff measure for the gauge h(s)=s^2 log log(1/s), with a deterministic constant. This is a natural extension of Le Gall's sphere theorem, and the boundary case is genuinely new: the boundary metric and gauge behave differently from the volume case, and the Bessel estimates are new. The half-plane corollary follows cleanly from the known coupling.

The core estimates look solid. Proposition 2.1 is a sharp occupation-time estimate for the 5D Bessel process, lifted to the bridge via absolute continuity (Corollary 2.2). Proposition 3.2 uses a dyadic covering argument with the upper bound on the boundary metric from Lemma 3.1; the Borel–Cantelli step works. Corollary 3.3 then gives the two-sided density bounds for the Hausdorff measure against Lebesgue measure. There are no fitted parameters and no circularity. The citation pattern is appropriate, mostly Le Gall's programme.

Now the soft spot. Theorem 3.4 aims to upgrade absolute continuity to a constant density. The proof writes psi'(0) = limsup_{eps->0} m_h^D([0,eps])/eps = kappa. The first equality is not justified. Proposition 1.7 only tells you that this limsup is a deterministic constant; it does not say the limit exists, and psi'(0) as a right derivative at the root need not exist without a Lebesgue point argument. Since the conclusion that the density is kappa a.e. rests on exactly this equality, this is a real gap—though, from what is in the paper, a fillable one. Replace psi'(0) by F(x) = limsup_{eps->0} m_h^D([x,x+eps])/eps. By Lemma 1.4 each F(x) has the same law as F(0)=kappa, so F(x)=kappa a.s. for each fixed x, hence a.e. x. At Lebesgue points of the density the limsup of the average equals the density, so the density is kappa a.e. That would complete the proof.

Minor annoyance: Corollary 3.3 has two limsup r->infinity that should be r->0. Also the reference to 'Theorem 3.4' in the reader's report corresponds to the paper's statement; the numbering is fine.

For whom: this is for the Brownian geometry community. It settles an open question and gives the right intrinsic characterisation of the boundary measure. I'd bring it to reading group. It deserves a serious referee; the main gap is local and repairable, not a conceptual failure. I'd send it out with a request to fix Theorem 3.4 along the lines above.

Referee Report

2 major / 6 minor

Summary. The paper proves that, on the boundary of the free Brownian disk, the uniform measure coincides almost surely with a constant multiple of the Hausdorff measure associated with the gauge function h(s)=s^2 log log(1/s), for a deterministic constant κ in (0,∞). The analogous statement for the Brownian half-plane is stated as a corollary. The proof strategy is to show that the boundary Hausdorff measure and Lebesgue measure on the boundary are mutually absolutely continuous with bounded densities, and then to use a rerooting-invariance argument to prove that the density is constant. The key estimates concern occupation times of five-dimensional Bessel processes and the resulting volume growth of metric balls centered on the boundary.

Significance. If correct, this is a natural and valuable result: it shows that the uniform measure on the boundary of the Brownian disk (and of the Brownian half-plane) is determined by the metric alone, extending Le Gall's analogous theorem for the Brownian sphere. The chosen gauge is not fitted ad hoc; it comes from the known occupation-time asymptotics of five-dimensional Brownian motion, and the constant κ is not built into the construction. The paper is well structured, and the main analytic estimates (Propositions 2.1, Corollary 2.2, Lemma 3.1, Proposition 3.2) are plausible and are connected by a clear chain of arguments. The main weakness is the final step in Theorem 3.4, where a pointwise derivative at the root is identified with a deterministic limsup without proof; this step is load-bearing but appears repairable by a standard limsup-density argument.

major comments (2)

[§3, Theorem 3.4] The proof contains a load-bearing identification that is not justified. The text first observes that the cumulative function ψ is differentiable λ-a.e. with derivative equal to the Radon–Nikodym derivative of m_h^D with respect to Lebesgue measure; this says nothing about the distinguished point 0. The displayed equality ψ′(0)=limsup_{ε↓0} m_h^D([0,ε])/ε is therefore asserted without support. Proposition 1.7 identifies the limsup as a deterministic constant κ, but it does not show that the right derivative at 0 exists or that the limsup equals a derivative. Since the subsequent rerooting argument transfers ψ′(0)=κ to every x and then identifies ψ′ with the density λ-a.e., the conclusion of Theorem 3.4 rests entirely on this unsupported equality. A fix is available in scope: define F(x)=limsup_{ε↓0} m_h^D([x,x+ε])/ε. By Lemma 1.4, F(x) has the same law as F(0) for each fixed x; Proposition 1.7 gives F(0)=κ a.s.; and Lebesgue's differentiation theorem gives F(x)=dm_h^D/dλ(x) for λ-a.e. x on a full-probability event. A Fubini argument then yields density κ λ-a.e. on a full-probability event. This replacement should be written out explicitly.
[§1.4, Proposition 1.7] The proof that the two limsup constants φ_D and φ_H are governed by the same measurable function Φ is only sketched. For the half-plane, the approximation of D∞ by the truncated distance D∞^{(η)} on small neighborhoods is argued, but the analogous reduction of φ_D to data in an arbitrarily small neighborhood of 0 in the disk is not given in the same detail. Since the conclusion that the two limsup constants coincide is needed both for Corollary 0.2 and for the deterministic nature of κ in Theorem 3.4, this step should be expanded into a complete argument.

minor comments (6)

[§1.4, after Eq. (12)] The gauge h(s)=s^2 log log(1/s) is not defined at s=1; one should specify that h is considered on a small interval [0,δ] with δ<e^{-e} so that log log(1/s) is well defined and nonnegative.
[§3, Corollary 3.3] In the lower-bound part of the proof, the expression "lim sup_{r→∞}" appears twice and should read "lim sup_{r↓0}" or "lim sup_{r↓0+}"; as written it is a typo that obscures the argument.
[§2, Lemma 3.1 proof] The Borel–Cantelli argument should explicitly sum over i∈{1,...,2^n} after the displayed probability bound; the displayed estimate is for a fixed i, and the union over i contributes a factor 2^n, which still gives a summable bound.
[§1.3, after Eq. (8)] The notation D∞ is used both for the pseudo-metric and for the resulting metric space (H,D∞); this is a mild abuse but should be flagged for readability.
[Introduction and abstract] There are several typographical and grammatical errors (e.g., "gau ge", "Mathémathiques", "Annal. Inst. H. Poincarré", "one can defined"). A careful proofreading pass is needed.
[§3, Theorem 3.4] The statement "ψ′(x) and ψ′(0) have the same law for every x" is used without defining ψ′(0) as a random variable. Even after defining it as a limsup, the equality-of-laws statement should be phrased in terms of the shifted boundary measure and should be followed by a Fubini argument rather than a pointwise conclusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the gauge is externally motivated and κ is not fitted; the only flagged issue in Theorem 3.4 is a technical gap, not circularity.

full rationale

The derivation is self-contained against external benchmarks. The gauge h(s)=s^2 log log(1/s) is imported from known occupation-time asymptotics for 5-dimensional Bessel processes (Ciesielski–Taylor [6] and Le Gall [14]), not fitted to the target uniform/Hausdorff ratio. The constant κ in Proposition 1.7 is defined as the almost-sure deterministic value of limsup_{ε↓0} m_h^D([0,ε])/ε, established by the zero–one law for the half-plane germ algebra; it is not defined as the ratio of the two measures whose equality is to be proved. Mutual absolute continuity (Corollary 3.3) plus rerooting invariance (Lemma 1.4) are then used to transfer the density value from the root to the whole boundary, producing m_h^D = κ λ. No load-bearing self-citation chain is present: the cited constructions of the Brownian disk/half-plane ([3],[5],[15],[11]) and the occupation-time estimates ([6],[14]) are external tools, and the author's own prior work is not invoked. One genuine technical issue is flagged: in the proof of Theorem 3.4, the equality "ψ′(0)=lim sup_{ε↓0+} m_h^D([0,ε])/ε=κ" is asserted without proving that ψ is right-differentiable at 0 or that the pointwise derivative equals the limsup; without such a justification the transfer from the root to a.e. point is incomplete. This is an omitted proof step, not a circularity: κ is not fitted from the conclusion, and the claimed equality does not make Theorem 0.1 equivalent to its inputs by construction. A fix would replace ψ′(0) by the translation-covariant limsup density, which is stationary by Lemma 1.4 and deterministic at 0 by Proposition 1.7. Hence the circularity score is 0.

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

The paper introduces no new entities or fitted parameters. It relies on standard constructions of Brownian geometry and classical estimates.

assumptions (5)

standard math The Brownian snake excursion measure N0 exists and satisfies the stated properties (Definition 1.2).
Standard construction in Brownian snake theory.
domain assumption The Brownian disk constructed from a 5D Bessel bridge and a Poisson snake process is the limit of Boltzmann triangulations in GHPU topology (Lemma 1.4, citing [1,3,10]).
Used to establish rerooting invariance.
standard math Lemma 1.6 (mass distribution criterion for Hausdorff measures) from Duquesne-Le Gall.
Core tool to compare measures.
standard math Blumenthal zero-one law for Bessel processes (used in Lemma 1.5).
Used to show the boundary density constant is deterministic.
domain assumption Ciesielski-Taylor result on the exact Hausdorff measure of Brownian paths (reference [6])
Provides the limsup bound on occupation times in Corollary 3.3.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The Hausdorff measure of the boundary of the Brownian disk." pith.science (2026). https://pith.science/paper/CAYQ2VF5

@misc{pith2026241116541,
  author       = {Pith},
  title        = {Pith review of: The Hausdorff measure of the boundary of the Brownian disk},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CAYQ2VF5}},
  note         = {Machine review of arXiv:2411.16541}
}

abstract

Consider the boundary $\partial \mathbb D$ of the Brownian disk $\mathbb D$ as a metric space by endowing it with the (restriction of the) metric of $\mathbb D$. We show that the uniform measure on $\partial \mathbb D$ coincides with the Hausdorff measure associated with the gauge function $h(s)=\kappa s^2\log\log(1/s)$ for some deterministic constant $\kappa>0$. We also state the analogous result for the boundary of the Brownian half-plane $\mathbb H$. This proves in particular that the uniform measure on the boundary of the Brownian disk (resp. the Brownian half-plane) is determined by the metric on $\mathbb D$ (resp. on $\mathbb H$).

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 14 canonical work pages

[1]

Albenque, N

M. Albenque, N. Holden and X. Sun. Scaling limit of large t riangulations of polygons. Electronic J. of Probab. 25 (2020) article no. 135, 43

work page 2020
[2]

E. Baur, G. Miermont, and G. Ray. Classiﬁcation of scalin g limits of uniform quadrangula- tions with a boundary. Ann. Probab. 47.6 (2019), 3397–3477. ISSN: 0091-1798,2168-894X. DOI: 10.1214/18-aop1316. URL: https://doi.org/10.1214/18-aop1316

doi:10.1214/18-aop1316 2019
[3]

Bettinelli and G

J. Bettinelli and G. Miermont. Compact Brownian surface s I. Brownian disks. Prob. The- ory Related Fields 167 (2017), 555–614

work page 2017
[4]

Burago, Y

D. Burago, Y. Burago, and S. Ivanov. A course in metric geometry . Vol. 33. Graduate Studies in Mathematics. American Mathematical Society, Pr ovidence, RI, 2001. ISBN: 0-8218-2129-6. DOI: 10.1090/gsm/033. URL: https://doi.org/10.1090/gsm/033

doi:10.1090/gsm/033 2001
[5]

Caraceni and N

A. Caraceni and N. Curien. Geometry of the Uniform Inﬁnit e Half-Planar Quadrangula- tion. Random Structures Algorithms 52 (2018), 454–494

work page 2018
[6]

Ciesielski and S

Z. Ciesielski and S. J. Taylor. First passage times and so journ times for Brownian motion in space and the exact Hausdorﬀ measure of the sample p ath. Trans. Amer. Math. Soc. 103 (1962), 434–450. ISSN: 0002-9947,1088-6850. DOI: 10.2307/1993838. URL: https://doi.org/10.2307/1993838. 15

doi:10.2307/1993838 1962
[7]

Curien and J.-F

N. Curien and J.-F. Le Gall. The hull process of the Browni an plane. Probab. Theory Related Fields 166 (2016), 187–231

work page 2016
[8]

T.Duquesne and J.-F. Le Gall. The Hausdorﬀ measure of sta ble trees. Alea 1 (2006), 393–415

work page 2006

Show all 17 references

[9]

R. K. Getoor. The Brownian escape process. Ann. Probab. 7 (1979), 864–867

1979
[10]

Gwynne and J

E. Gwynne and J. Miller. Convergence of the free Boltzma nn quadrangulation with simple boundary to the Brownian disk. Ann. Inst. Henri Poincaré Probab. Stat. 55.1 (2019), 551–

2019
[11]

Gwynne and J

E. Gwynne and J. Miller. Scaling limit of the uniform inﬁ nite half-plane quadrangulation in the Gromov–Hausdorﬀ–Prokhorov–uniform topology. Electron. J. Probab. 22 (2017), Paper No. 84, 47. DOI: 10.1214/18-aihp891. URL: https://doi.org/10.1214/18-aihp891

2017 doi
[12]

J.-F. Le Gall. Brownian disks and the Brownian snake. Annal. Inst. H. Poincarré Probab. Stat. 55 (2019), 237–313

2019
[13]

J.-F. Le Gall. Geodesics in large planar maps and in the B rownian map. Acta Mathematica 205 (2010), 287–360

2010
[14]

J.-F. Le Gall. Sur la mesure de Hausdorﬀ de la courbe brow nienne. Séminaire de proba- bilités, XIX, 1983/84. Vol. 1123. Lecture Notes in Math. Springer, B erlin, 1985, 297–313

1983
[15]

J.-F. Le Gall. The Brownian disk viewed from a boundary p oint. Annal. Inst. H. Poincarré Probab. Stat. 58 (2022), 1091–1119

2022
[16]

J.-F. Le Gall. The volume measure of the Brownian sphere is a Hausdorﬀ measure. Elec- tronic J. of Probab. 27 (2022), article no. 113, 1–28. 16

2022
[589]

ISSN: 0246-0203,1778-7017

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Albenque, N

M. Albenque, N. Holden and X. Sun. Scaling limit of large t riangulations of polygons. Electronic J. of Probab. 25 (2020) article no. 135, 43

work page 2020

[2] [2]

E. Baur, G. Miermont, and G. Ray. Classiﬁcation of scalin g limits of uniform quadrangula- tions with a boundary. Ann. Probab. 47.6 (2019), 3397–3477. ISSN: 0091-1798,2168-894X. DOI: 10.1214/18-aop1316. URL: https://doi.org/10.1214/18-aop1316

doi:10.1214/18-aop1316 2019

[3] [3]

Bettinelli and G

J. Bettinelli and G. Miermont. Compact Brownian surface s I. Brownian disks. Prob. The- ory Related Fields 167 (2017), 555–614

work page 2017

[4] [4]

Burago, Y

D. Burago, Y. Burago, and S. Ivanov. A course in metric geometry . Vol. 33. Graduate Studies in Mathematics. American Mathematical Society, Pr ovidence, RI, 2001. ISBN: 0-8218-2129-6. DOI: 10.1090/gsm/033. URL: https://doi.org/10.1090/gsm/033

doi:10.1090/gsm/033 2001

[5] [5]

Caraceni and N

A. Caraceni and N. Curien. Geometry of the Uniform Inﬁnit e Half-Planar Quadrangula- tion. Random Structures Algorithms 52 (2018), 454–494

work page 2018

[6] [6]

Ciesielski and S

Z. Ciesielski and S. J. Taylor. First passage times and so journ times for Brownian motion in space and the exact Hausdorﬀ measure of the sample p ath. Trans. Amer. Math. Soc. 103 (1962), 434–450. ISSN: 0002-9947,1088-6850. DOI: 10.2307/1993838. URL: https://doi.org/10.2307/1993838. 15

doi:10.2307/1993838 1962

[7] [7]

Curien and J.-F

N. Curien and J.-F. Le Gall. The hull process of the Browni an plane. Probab. Theory Related Fields 166 (2016), 187–231

work page 2016

[8] [8]

T.Duquesne and J.-F. Le Gall. The Hausdorﬀ measure of sta ble trees. Alea 1 (2006), 393–415

work page 2006