Trace inequality for $BV(\Omega)$ in smooth domains

Jos\'e Claudio Sabina de Lis

Trace inequality for $BV(\Omega)$ in smooth domains

T0 review · 0 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 On bounded $C^{1,1}$ domains, the BV trace inequality holds with the sharp constant 1.

desk verdict Sharp BV trace inequality with constant 1 on C^{1,1} domains, with matching counterexamples; the main result is correct and the regularity threshold is essentially optimal. read the letter →

arxiv 2411.17325 v2 pith:YUJZNXGQ submitted 2024-11-26 math.AP

José Claudio Sabina de Lis This is my paper · ORCID

classification math.AP MSC 46E3526D1035J92

keywords traceinequalityfunctionsofboundedvariationsetsfiniteperimeterC^{11}domainpositivereachlowersemicontinuityone-Laplacianboundary

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 boundary trace inequality for functions of bounded variation, $\int_{\partial\Omega}|u| \le C_1\int_\Omega|Du| + C_2\int_\Omega|u|$, holds on bounded $C^{1,1}$ domains with the sharp coefficient $C_1 = 1$, and that this is the best possible value. The regularity assumption is essentially optimal: for every $0<\alpha<1$ there are $C^{1,\alpha}$ domains where the same inequality with $C_1=1$ fails for every finite $C_2$. The result matters because variational problems for the one-Laplacian and for phase transitions with boundary contact terms need exactly this constant-one inequality to preserve lower semicontinuity of the energy. The paper also identifies the precise regularity hidden behind a widely used 'smooth domain' assumption in an earlier semicontinuity theorem.

What carries the argument

The load-bearing object is the tubular coordinate system built from the inward unit normal field $n$ on $\partial\Omega$. On a $C^{1,1}$ boundary the normal is Lipschitz, so the charts $h_i(s,t)=g_i(s)+t n(g_i(s))$ are bi-Lipschitz onto a uniform neighborhood, the Jacobian $J(s,t)=\det Dh_i(s,t)$ stays positive, and the ratio $|\partial_t J|/J$ is bounded by a constant related to the mean curvature of the boundary, which is exactly the positive-reach property. The trace inequality is obtained by writing the surface integral as $\int |u_i(s,0)|J(s,0)\,ds$, using the identity $\partial_t(u_iJ)=\partial_tu_i J+u_i\partial_tJ$ to trade the boundary term for volume integrals, and absorbing $|\partial_tJ|/J$ into the $\int_\Omega|u|$ term. The boundedness of this ratio is precisely where the argument uses positive reach; without a Lipschitz normal the local estimate breaks, and the $C^{1,\alpha}$ counterexamples exhibit that breakdown.

What would settle it

Take the $C^{1,\alpha}$ domain of Example 9 and the sets $E_n=\{x_N>|x'|^{1+\alpha}/(1+\alpha),\,x_N<\psi(r_n)\}$ with $r_n\to0$. If the claimed optimal inequality held there, then $\int_{\partial\Omega}|u_n|-\int_\Omega|Du_n|$ would have to be bounded by $C\int_\Omega|u_n|$; the paper's asymptotic gives the left side of order $r_n^{N-1+2\alpha}$ and the right side of order $r_n^{N+\alpha}$, so the ratio tends to infinity because $\alpha<1$. Recomputing that expansion is the direct check.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is Theorem 7: if $\Omega\subset\mathbb{R}^N$ is a bounded $C^{1,1}$ domain, then there exists a constant $C=C(\Omega)$ such that $\int_{\partial\Omega}|u| \le \int_\Omega|Du| + C\int_\Omega|u|$ for every $u\in BV(\Omega)$. The coefficient of the total variation is exactly $1$, and Lemma 3 shows that $C_1\ge 1$ for any competing inequality, so the constant is optimal. The proof decomposes $u$ with a partition of unity, uses the normal-coordinate charts $h_i(s,t)=g_i(s)+t n(g_i(s))$, and converts the boundary integral into interior integrals through $\partial_t(u_iJ)=\partial_tu_iJ+u_i\partial_tJ$, with boundedness of $|\partial_tJ|/J$ supplied by the Lipschitz normal field. Example 9 then shows optimality by constructing $C^{1,\alpha}$ cusp domains, with $0<\alpha<1$, where the characteristic function of a thin layer has $\int_{\partial\Omega}|u_n|-\int_\Omega|Du_n|$ of order $r_n^{N-1+2\alpha}$ while $\int_\Omega|u_n|$ is only of order $r_n^{N+\alpha}$, so no finite $C$ can close the gap. Example 10 transfers this failure to the lower semicontinuity of $J_-(u)=\int_\Omega|Du|-\int_{\partial\Omega}|u|$.

Load-bearing premise

The proof collapses if the boundary's unit normal field is not Lipschitz, because then the normal-coordinate map has no uniform positive Jacobian with bounded logarithmic derivative; the $C^{1,\alpha}$ examples show this is not a removable technicality.

Editorial extensions

If this is right

On every bounded $C^{1,1}$ domain, the trace inequality holds with $C_1=1$; since $C_1\ge1$ is forced, this is the optimal constant for all such domains.
The lower semicontinuity of $J(u)=\int_\Omega|Du|+\int_{\partial\Omega}F(x,u)$ for functions $F$ with Lipschitz constant at most $1$ in $u$ is valid on $C^{1,1}$ domains, and Example 10 shows that $C^{1,\alpha}$ regularity does not suffice.
The passage from $W^{1,1}$ to $BV$ by strict approximation preserves the inequality with the same constants, so the result applies to characteristic functions of sets of finite perimeter.
On merely $C^1$ domains the inequality holds with $C_1$ arbitrarily close to $1$ but not necessarily equal to $1$, and on piecewise-$C^1$ domains $C_1$ can be forced to exceed $1$.

Reading between the lines

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

Beyond the paper's own statement, the proof's mechanism suggests that the same $C_1=1$ result should extend to any bounded domain whose boundary has positive reach, including some piecewise-smooth boundaries with no reentrant corners.
For the one-Laplacian and phase-transition literature that invokes lower semicontinuity under a vague 'smooth domain' hypothesis, the paper's sharpened threshold suggests those arguments should either assume $C^{1,1}$ or prove the inequality directly for their specific domains.
A quantitative version of Theorem 7 is likely available: the constant $C$ can be chosen in terms of the supremum of $|(N-1)H|$ and the chart data, so domains in a family with uniformly controlled mean curvature would satisfy the inequality with a uniform constant.

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

Short version: this paper closes a real gap. Modica assumed C1=1 for smooth domains; Giusti only suggested it; Motron claimed piecewise C^1 and is wrong. Sabina de Lis proves the constant-1 inequality for bounded C^{1,1} domains and gives explicit C^{1,α} domains, 0<α<1, where the inequality fails. The examples are simple and convincing; they also show lower semicontinuity of J(u)=∫Ω|Du| + ∫∂Ω F(x,u) fails for C^{1,α}, so the C^{1,1} assumption is not a technical artifact. This is a genuine within-field advance, not a big conceptual leap.

The proof is honest local geometry plus partition of unity. The chart h_i(s,t)=g_i(s)+t n(s) with J(s,t)=det Dh_i is standard; for C^{1,1} boundary, n is Lipschitz, J(s,t) is polynomial in t with L∞ coefficients, and J(s,0) is bounded below on compact chart domains, so the required uniform estimates on J and |∂tJ|/J hold. The C^2 case is presented first and then extended by Rademacher; that is the right order. The correction of Motron's piecewise C^1 claim is backed by the cone example showing C1≥√(1+L^2)>1. The author also computes the normal curvature blow-up for the C^{1,α} example, which explains why the positive-reach machinery fails there.

Soft spots are minor and mostly implicit. The proof would be easier to check if the uniform lower bound on J(s,t) and the finiteness of C_i=sup|∂tJ|/J were stated explicitly rather than left in the phrase "J(s,t) can be assumed positive for ε small." Similarly, the change-of-variables formula for Lipschitz homeomorphisms is cited but the details that h_i is bi-Lipschitz with Lipschitz inverse are only in Theorem 2. A referee should ask for those two sentences. The paper is a bit dense in Section 3; the polynomial-in-t observation for J deserves a display. None of this threatens the argument.

Citation pattern is fine: Modica, Giusti, and Motron are discussed accurately, and the self-citations are for computational details and do not carry the proof.

Bottom line: the paper deserves a serious referee. I would send it out. The results are right, the counterexamples are explicit, and the sharpness at C^{1,1} versus C^{1,α} is exactly the kind of precision people working on 1-Laplacian boundary value problems need. I would not insist on major changes.

Referee Report

0 major / 6 minor

Summary. The paper studies the trace inequality for BV functions on bounded domains, aiming at the best constant C1=1 in (1.1). The main result (Theorem 7) establishes that for every bounded C^{1,1} domain Ω there exists C(Ω) such that ∫∂Ω|u| ≤ ∫Ω|Du| + C∫Ω|u| for all u in BV(Ω). The proof uses a tubular neighborhood coordinate system based on the signed distance function, a partition of unity, and a first-order computation showing that the Jacobian J(s,t) satisfies the needed a priori estimate. The paper also proves that for C^1 domains the constant C1 can be taken arbitrarily close to 1 (Theorem 4), shows that C1 cannot be below 1 (Lemma 3), and constructs explicit examples showing that C^{1,α} regularity with 0<α<1 does not suffice (Example 9) and that the associated lower semicontinuity property fails in that setting (Example 10).

Significance. If accepted, the paper resolves a natural question on the exact constant in the BV trace inequality for smooth domains, showing C1=1 for C^{1,1} domains and providing elementary counterexamples for lower Hölder regularity. The paper is rigorous and self-contained: the proof of Theorem 7 is based on explicit local computations, the dependence of the constant on the charts and the mean curvature is made explicit, and the lower-bound examples are explicit and verifiable. The link between positivity of the reach and C^{1,1} regularity (Theorem 1) is valuable, and the paper clarifies the implicit use of C1=1 in Modica's semicontinuity theorem. The examples are carefully chosen to separate the C^{1,1} threshold from C^{1,α}.

minor comments (6)

[Section 2.2 / Section 3] In the proof of Theorem 7 for the C^{1,1} case, the quantity C_i = sup_{U_i×(0,ε)} |∂tJ|/J is used; this is finite only if the chart domains are chosen to be bounded and the Jacobian J(s,0) is uniformly positive on the support of the partition of unity. Since ∂Ω is compact, such a finite sub-atlas exists, but this should be stated explicitly, otherwise the supremum may be infinite for a chart with unbounded parameter domain.
[Equation (2.4)] In the definition of h_i(s,t) = g_i(s) + t n(g(si)), the argument of n should be g_i(s), not g(si); the current notation is a typo.
[Equation (3.18)] After (3.18), the sentence 'the change of variables x = h_i(x,t) must be performed' should read x = h_i(s,t).
[Introduction / Section 4] The introduction's phrase 'optimum smoothness requirements' suggests a necessary-and-sufficient characterization, whereas the paper actually establishes sufficiency of C^{1,1} and non-sufficiency of the class C^{1,α} for every α<1; it does not rule out that some individual C^{1,α} domains satisfy the inequality. The wording should be softened to avoid overclaiming.
[Section 4, Example 9] In the definition of E_n, the sequence r_n is introduced without an explicit monotonicity or convergence statement; for clarity, state that r_n → 0+ as n → ∞.
[Declarations] The sentence 'founded by MCIN/AEI' should read 'funded by MCIN/AEI'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the trace inequality is derived from a direct Jacobian estimate, and no fitted quantity or self-citation carries the proof.

full rationale

The central claim, Theorem 7, is proved by a self-contained local computation. In each boundary chart h_i(s,t)=g_i(s)+t n(g_i(s)), the paper derives estimate (3.18) directly from the identity (3.16), with the coefficient of ∫Ω|Du| being exactly 1 on every chart; the constant C_i depends only on the geometric quantity sup |∂tJ|/J. No parameter is fitted to make the desired inequality hold. The lower bound C1≥1 in Lemma 3 is obtained by an independent perimeter-approximation argument, and the counterexamples in Examples 8 and 9 are explicit constructions, not consequences of the theorem. The cited works [12], [22], and [21] are used for background or as context for the semicontinuity functional; the proof of Theorem 7 does not rely on the conclusions it is trying to establish. The self-citations [24] and [25] appear only in a computation-detail remark and in a list of applications, respectively, and neither is load-bearing. There is no self-definitional step, no fitted input renamed as a prediction, no imported uniqueness theorem, and no ansatz smuggled in by citation. The derivation chain from the Jacobian estimate (3.18) to the trace inequality (3.11) is complete and independent of the examples that establish sharpness.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central derivation introduces no fitted constants and no ad hoc entities. The proof rests on standard geometric measure and PDE background, listed above, plus the C^{1,1} hypothesis. Constants C_i are geometric suprema shown finite, not fitted parameters. The examples use parametrized domains only as counterexamples.

assumptions (6)

standard math Existence and strict-continuity of the BV trace operator on Lipschitz domains (Ambrosio-Fusco-Pallara).
Used in Section 2.1 and in Theorem 7 to pass from C^1 approximations to BV functions.
standard math Meyer-Serrin approximation theorem for BV (Evans-Gariepy Th. 2.2.2).
Used in Theorem 7 proof to approximate arbitrary BV functions by smooth functions in the strict topology.
standard math For C^1 boundaries, positive reach r(y)>0 for all y is equivalent to C^{1,1} regularity (Federer, Lucas; Theorem 1).
Justifies that C^{1,1} supplies the Lipschitz tubular coordinates used in the proof.
standard math Tubular neighborhood properties: projection π is Lipschitz, h_i is a Lipschitz homeomorphism with Lipschitz inverse, signed distance is C^1 (Theorem 2, [9], [15]).
These facts allow the chart decomposition and the change of variables in Section 3.
standard math Change-of-variables formula for bi-Lipschitz maps (Evans-Gariepy Th. 3.3.2).
Used to transform integrals over (s,t) into integrals over Ω for the Lipschitz charts in the C^{1,1} case.
domain assumption Ω is a bounded C^{1,1} domain.
The hypothesis of Theorem 7; it provides the Lipschitz normal field and finite mean-curvature-type bounds needed for C1=1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Trace inequality for $BV(\Omega)$ in smooth domains." pith.science (2026). https://pith.science/paper/YUJZNXGQ

@misc{pith2026241117325,
  author       = {Pith},
  title        = {Pith review of: Trace inequality for $BV(\Omega)$ in smooth domains},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YUJZNXGQ}},
  note         = {Machine review of arXiv:2411.17325}
}

abstract

This work is devoted to prove an optimum version of the trace inequality associated to the embedding $BV(\Omega)\subset L^1(\partial\Omega)$. Special emphasis is placed on the regularity that the domain $\Omega$ should exhibit for this result to be valid.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 27 canonical work pages

[1]

R. A. Adams. Sobolev spaces. Pure and Applied Mathematics, Vol. 65. Aca- demic Press [Harcourt Brace Jovanovich, Publishers], New York-Lo ndon, 1975

work page 1975
[2]

Ambrosio, N

L. Ambrosio, N. Fusco, and D. Pallara. Functions of Bounded Variation and Free Discontinuity Problems. Oxford Mathematical Monographs. The Claren- don Press, Oxford University Press, New York, 2000

work page 2000
[3]

Andreu, J

F. Andreu, J. M. Maz´ on, and J. S. Moll. The total variation ﬂow w ith nonlinear boundary conditions. Asymptot. Anal. , 43(1-2):9–46, 2005

work page 2005
[4]

Anzellotti and M

G. Anzellotti and M. Giaquinta. BV functions and traces. Rend. Sem. Mat. Univ. Padova , 60:1–21 (1979), 1978

work page 1979
[5]

Barbato, F

R. Barbato, F. Della Pietra, and G. Piscitelli. On the ﬁrst Robin eigen value of the Finsler p-Laplace operator as p → 1. J. Math. Anal. Appl. , 540(2):Paper No. 128660, 25, 2024

work page 2024
[6]

Della Pietra, C

F. Della Pietra, C. Nitsch, F. Oliva, and C. Trombetti. On the behav ior of the ﬁrst eigenvalue of the p-Laplacian with Robin boundary conditions as p goes to 1. Adv. Calc. Var. , 16(4):1123–1135, 2023

work page 2023
[7]

Della Pietra, F

F. Della Pietra, F. Oliva, and S. Segura de Le´ on. Behaviour of solu tions to p-Laplacian with Robin boundary conditions as p goes to 1. Proc. Roy. Soc. Edinburgh Sect. A , 154(1):105–130, 2024

work page 2024
[8]

L. C. Evans and R. F. Gariepy. Measure theory and ﬁne properties of functions . Textbooks in Mathematics. CRC Press, Boca Raton, FL, revised ed ition, 2015

work page 2015

Show all 27 references

[9]

H. Federer. Curvature measures. Trans. Amer. Math. Soc. , 93:418–491, 1959

1959
[10]

Gagliardo

E. Gagliardo. Caratterizzazioni delle tracce sulla frontiera rela tive ad alcune classi di funzioni in n variabili. Rend. Sem. Mat. Univ. Padova , 27:284–305, 1957

1957
[11]

Gilbarg and N

D. Gilbarg and N. S. Trudinger. Elliptic partial diﬀerential equations of sec- ond order , volume 224 of Grundlehren der mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences] . Springer-Verlag, Berlin, second edition, 1983

1983
[12]

E. Giusti. The equilibrium conﬁguration of liquid drops. J. Reine Angew. Math., 321:53–63, 1981

1981
[13]

Guillemin and A

V. Guillemin and A. Pollack. Diﬀerential topology . Prentice-Hall, Inc., Engle- wood Cliﬀs, NJ, 1974. TRACE INEQUALITY IN BV 21

1974
[14]

Hauer and J

D. Hauer and J. M. Maz´ on. The Dirichlet-to-Neumann operato r associated with the 1-Laplacian and evolution problems. Calc. Var. Partial Diﬀerential Equations, 61(1):Paper No. 37, 50, 2022

2022
[15]

S. G. Krantz and H. R. Parks. Distance to Ck hypersurfaces. J. Diﬀerential Equations, 40(1):116–120, 1981

1981
[16]

Latorre and S

M. Latorre and S. Segura de Le´ on. Existence and comparison results for an elliptic equation involving the 1-Laplacian and L1-data. J. Evol. Equ. , 18(1):1– 28, 2018

2018
[17]

Lewicka and Y

M. Lewicka and Y. Peres. Which domains have two-sided support ing unit spheres at every boundary point? Expo. Math., 38(4):548–558, 2020

2020
[18]

K. R. Lucas. Submanifolds of dimension (n − 1) in E n with normals satisfying a Lipschitz condition . ProQuest LLC, Ann Arbor, MI, 1957. Thesis (Ph.D.)– University of Kansas

1957
[19]

Massari and L

U. Massari and L. Pepe. Sull’approssimazione degli aperti lipsch itziani di Rn con variet` a diﬀerenziabili. Boll. Un. Mat. Ital. (4) , 10:532–544, 1974

1974
[20]

J. M. Maz´ on, J. D. Rossi, and S. Segura de Le´ on. The 1-Lapla cian elliptic equation with inhomogeneous Robin boundary conditions. Diﬀerential Integral Equations, 28(5-6):409–430, 2015

2015
[21]

L. Modica. Gradient theory of phase transitions with boundary contact energy. Ann. Inst. H. Poincar´ e Anal. Non Lin´ eaire, 4(5):487–512, 1987

1987
[22]

M. Motron. Around the best constants for the Sobolev trace map from W 1,1(Ω) into L1(∂Ω). Asymptot. Anal. , 29(1):69–90, 2002

2002
[23]

J. Neˇ cas. Direct methods in the theory of elliptic equations . Springer Mono- graphs in Mathematics. Springer, Heidelberg, 2012. Translated fr om the 1967 French original by Gerard Tronel and Alois Kufner, Editorial coord ination and preface by ˇS´ arka Neˇ casov´ a and a co...

2012
[24]

Pardo, A

R. Pardo, A. L. Pereira, and J. C. Sabina de Lis. The tangential variation of a localized ﬂux-type eigenvalue problem. J. Diﬀerential Equations , 252(3):2104– 2130, 2012

2012
[25]

J. C. Sabina de Lis and S. Segura de Le´ on. Higher Robin eigenvalu es for the p-Laplacian operator as p approaches 1. Calc. Var. Partial Diﬀerential Equa- tions, 63(7):Paper No. 169, 2024

2024
[26]

L. Tartar. An introduction to Sobolev spaces and interpolation spaces , volume 3 of Lecture Notes of the Unione Matematica Italiana . Springer, Berlin; UMI, Bologna, 2007

2007
[27]

J. A. Thorpe. Elementary topics in diﬀerential geometry . Undergraduate Texts in Mathematics. Springer-Verlag, New York, 1994. Corrected rep rint of the 1979 original. Jos´ e Sabina de Lis Departamento de An ´alisis Matem ´atico and IUEA, Universidad de La Laguna, P. O. Box 45...

1994

Pith tools

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

[1] [1]

R. A. Adams. Sobolev spaces. Pure and Applied Mathematics, Vol. 65. Aca- demic Press [Harcourt Brace Jovanovich, Publishers], New York-Lo ndon, 1975

work page 1975

[2] [2]

Ambrosio, N

L. Ambrosio, N. Fusco, and D. Pallara. Functions of Bounded Variation and Free Discontinuity Problems. Oxford Mathematical Monographs. The Claren- don Press, Oxford University Press, New York, 2000

work page 2000

[3] [3]

Andreu, J

F. Andreu, J. M. Maz´ on, and J. S. Moll. The total variation ﬂow w ith nonlinear boundary conditions. Asymptot. Anal. , 43(1-2):9–46, 2005

work page 2005

[4] [4]

Anzellotti and M

G. Anzellotti and M. Giaquinta. BV functions and traces. Rend. Sem. Mat. Univ. Padova , 60:1–21 (1979), 1978

work page 1979

[5] [5]

Barbato, F

R. Barbato, F. Della Pietra, and G. Piscitelli. On the ﬁrst Robin eigen value of the Finsler p-Laplace operator as p → 1. J. Math. Anal. Appl. , 540(2):Paper No. 128660, 25, 2024

work page 2024

[6] [6]

Della Pietra, C

F. Della Pietra, C. Nitsch, F. Oliva, and C. Trombetti. On the behav ior of the ﬁrst eigenvalue of the p-Laplacian with Robin boundary conditions as p goes to 1. Adv. Calc. Var. , 16(4):1123–1135, 2023

work page 2023

[7] [7]

Della Pietra, F

F. Della Pietra, F. Oliva, and S. Segura de Le´ on. Behaviour of solu tions to p-Laplacian with Robin boundary conditions as p goes to 1. Proc. Roy. Soc. Edinburgh Sect. A , 154(1):105–130, 2024

work page 2024

[8] [8]

L. C. Evans and R. F. Gariepy. Measure theory and ﬁne properties of functions . Textbooks in Mathematics. CRC Press, Boca Raton, FL, revised ed ition, 2015

work page 2015