Elliptic regularity estimates with optimized constants and applications

Boyan Sirakov; Philippe Souplet

REVIEW 1 major objections 4 minor 56 references

Elliptic regularity estimates with optimized constants and applications

T0 review · 1 major / 4 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 This paper proves optimized constants for the main elliptic regularity estimates and supplies counterexamples showing none can be improved.

desk verdict Sharp-constant elliptic estimates, mostly solid and honest, with one load-bearing imported Harnack iteration that a referee must verify. read the letter →

arxiv 2411.19367 v2 pith:YGBFWLCK submitted 2024-11-28 math.AP

Boyan Sirakov , Philippe Souplet This is my paper · ORCID

classification math.AP MSC 35B4535B5035B6535J1535P15

keywords optimizedconstantsellipticregularityC^{1α}estimatesHarnackinequalityHopf–OleiniklemmaLandisconjecturefirsteigenvalueMorel–Oswald

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

Classical a priori estimates for uniformly elliptic equations in divergence form carry constants that depend on the coefficients and the size of the domain, but the classical proofs leave that dependence far from sharp. This paper establishes versions of the generalized maximum principle, the $C^{{1,α}}$ estimate, the quantitative Hopf lemma, the boundary weak Harnack inequality, and the logarithmic gradient estimate in which the constant is optimized: the exponential growth is linear in the coefficient measure M and the geodesic diameter D of the domain. It also shows by counterexamples that no slower growth in M, R, or D is possible for the drift and potential coefficients b2 and c, and it uses the optimized estimates to prove a Landis-type lower bound on ∫_{∂B_R}|u| and upper and lower bounds for the first eigenvalue. A sympathetic reader should care because these are the constants that matter whenever the estimates are used to prove existence, uniqueness, decay, or spectral statements.

What carries the argument

The machine is the optimized boundary weak Harnack inequality in Theorem 5.1, combined with a duality argument using the adjoint operator, its positive first eigenfunction, and an integration-by-parts lemma. The quantity M in (1.7) collects uniformly local norms of A, b1, b2, c raised to exponents βq and γq, together with a radius r0 in (3.2) chosen so that after rescaling each local operator has uniformly bounded coefficients; this is what converts classical non-optimized estimates into estimates with linear-in-M exponents. A Harnack chain whose length is controlled by the geodesic diameter D carries the estimates across the domain, and optimality is shown by explicit radial or eigenfunction-based solutions.

What would settle it

Compute the ratio in (2.2) for the explicit radial solutions in Propositions 4.1 and 4.3 (for example u = $e^{{λ(1−|x|^2)/2}}$−1 with operators −Δ+c and −Δ+b·∇) as λ grows: if the ratio sup v/d divided by $e^{{C0 M R}}$ $R^{{1−n/q}}$||f||_q is unbounded, the exponential constant is wrong; alternatively, run the Section 12.3 iteration on the half-ball with b=λx and c=0 to check whether the weak Harnack constant grows like $e^{{Cλ}}$ rather than $e^{{Cλ^2}}$.

Watch

Extended reading notes

Core claim

The paper's central claim is that, for operators L satisfying (1.3)-(1.4) on bounded $C^{{1,ᾱ}}$ domains, the sharp a priori estimates take the forms stated in Theorems 2.1–2.9 and 3.1–3.4. In particular, sup_{B_R} v/d ≤ C0 $e^{{C0 M R}}$ $R^{{1−n/q}}$ ||f^+||_{L^q(B_R)} holds under div(b1)+c ≤ 0, with M defined in (1.7) through uniformly local coefficient norms and a coefficient-adapted radius r0; the quantitative Hopf estimate u(x) ≥ $e^{{−C0(1+MR)}}$ $R^{{−n}}$ ||f||_{$L^{1}$_d} d(x) and the log-gradient estimate d|∇u|/u ≤ C0 max{1,Md} also hold with linear-in-M exponents. The paper proves these bounds and, in Propositions 4.1–4.4, constructs sequences of operators with growing b2 or c for which the exponential in M is attained, so the dependence cannot be improved.

Load-bearing premise

The load-bearing premise is the optimized boundary weak Harnack inequality (Theorem 5.1), whose proof imports an iteration argument from earlier results rather than re-deriving it in full; if that imported constant were larger than claimed, the linear-in-M exponents in the main estimates would degrade.

Editorial extensions

If this is right

The Landis lower estimate becomes ∫_{∂B_R}|u| dσ ≥ e^{−C0 K R} ∫_{B_R}|u| for any weak solution of Lu = 0 in R^n with λ1(−L,R^n) ≥ 0, where K is the explicit coefficient data in (2.14).
The first eigenvalue satisfies λ1(−L,B_R) ≤ C0(M+R^{-1})^2, and if λ1(−L,B_{R+κ}) ≥ 0 then λ1(−L,B_R) ≥ e^{−C0(M+κ^{-1})R} R^{−2}.
The quantitative Hopf–Oleinik inequality holds for general divergence operators: u(x) ≥ e^{−C0(1+MR)} R^{−n} ||f||_{L^1_d(B_R)} d(x).
For positive solutions of homogeneous equations, d|∇u|/u ≤ C0 max{1,Md}, and for the inhomogeneous equation with f ≥ 0 an additional term involving ||f||_{L^q}/||f||_{L^1_d} appears.
No estimate with slower growth in the drift or potential coefficients is possible: Propositions 4.1–4.4 exhibit operators attaining the exponential factor in M.

Reading between the lines

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

Beyond the paper: the same coefficient-adapted rescaling that powers the proofs should carry the optimized constants to operators with VMO or merely continuous leading coefficients near the boundary, since only uniformly local norms enter M; the Hölder assumptions in (1.4) feed the iteration but may not be the true threshold.
Beyond the paper: the quantitative Hopf bound with the L^1_d norm is built from a duality with the first eigenfunction of the adjoint, so the same argument should yield Green-function lower bounds with explicit constants for operators with first-order terms, which currently are only known in special cases.
Beyond the paper: the n=1 inhomogeneous log-gradient bound suggests that the failure of the estimate for n≥2 and sign-changing f is a dimension effect; a finite-difference check on the explicit example in (11.16) would test whether a positive-part condition on f suffices in higher dimensions.

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

Bottom line: this is a serious paper and the main claims look right, but the proof of Theorem 5.1 is where I would send a referee before believing the headline constants. The new material is real: the duality argument for the Lipschitz bound with exponent linear in M, the first optimized Morel-Oswald/Hopf estimate for general divergence operators, the nonhomogeneous log-gradient estimate, and the counterexamples that match the claimed exponential rates. The authors are also unusually straight about what is open: Remark 2.2 on the unoptimized case, partial optimality of (2.19), the necessity of the spectral positivity assumption.

I checked the one delicate computation the reader flagged, the adjoint-factorization step in Lemma 8.1 Step 3, and it is consistent: the drift coefficient and the identity div(b1hat)+chat = L* phi = 0 check out. The estimates are scale invariant and the counterexamples use the actual M of the constructed operators, so no parameter fitting is hidden in the optimality claims.

The soft spot is Theorem 5.1, the optimized global boundary weak Harnack inequality. It is the load-bearing result: Theorems 2.8, 2.9, 3.3, and 3.4 all pass through it, and its proof in Section 12.3 is not self-contained. The local estimates (12.31)-(12.33) are verified, but the Harnack-chain iteration is imported from [54] with "we can repeat the iteration argument on [54, p.12]". If that iteration gives a C0^N factor, then with the N bound from Proposition 12.2 the stated linear-exponential form follows. If it gave something like C0^{N^2}, the main estimates would lose their claimed linear-in-M form. I have no reason to think the imported constant is wrong, since these are the authors' own published results, but it is a genuine open check, not a cosmetic gap. A referee should work through that chain.

Also the abstract's optimality claim overstates slightly: the counterexamples cover the b2 and c dependencies, and the nonhomogeneous log-gradient estimate is only partly shown optimal. That is disclosed in the text, so it is a presentation issue, not a substantive flaw.

Who this is for: specialists in elliptic regularity who care about sharp constants, and people working on Landis-type quantitative unique continuation. It deserves a serious referee. I would send it to review, and I would want the referee to focus on Section 12.3 rather than the main theorems.

Referee Report

1 major / 4 minor

Summary. The paper revisits classical linear second-order uniformly elliptic equations in divergence form on bounded C^{1,ᾱ} domains, Lu = div(A Du + b1 u) + b2·Du + c u, and proves versions of the generalized maximum principle, Lipschitz/C^{1,α} estimates, the Hopf-Oleinik lemma, boundary weak Harnack inequalities, and logarithmic gradient estimates in which the constants depend optimally on a uniformly local, scale-invariant norm M of the coefficients and on the geodesic diameter D. The main estimates have the form e^{C0 M D} (or variants) with linear dependence on M and D, which improves on previously available superlinear dependence for the maximum principle. The paper also proves upper and lower bounds for the first eigenvalue, gives Landis-conjecture-type lower bounds for solutions in balls, and shows by explicit counterexamples that the exponential dependence on M is optimal for the main estimates. The proofs combine rescalings, Harnack chains, duality arguments, and a doubling/contradiction method, with detailed arguments in Sections 6–11 and auxiliary results in Section 12.

Significance. If correct, the results settle the sharp dependence of several classical a priori estimates on coefficient norms and domain size. The optimized Stampacchia–Trudinger estimate with exponent linear in M, the new up-to-the-boundary log-gradient bound for equations with right-hand side, and the optimized quantitative Hopf lemma are significant advances over the existing literature, and the counterexamples show that no better growth in M is possible. The paper is also careful about its own limitations: Remark 2.2 leaves an open problem, Remark 4.2 flags that the inhomogeneous log-gradient term is not known to be optimal, and Proposition 2.7 shows that the spectral assumption in the Landis-type result is necessary rather than merely technical. The authors ship detailed proofs, not just sketches, and the constructions in Propositions 4.1–4.4 use the actual quantities M of the operators involved, so no fitted parameter is disguised as a prediction.

major comments (1)

[Section 12.3, Theorem 5.1] The proof of the optimized global weak Harnack inequality, Theorem 5.1, is not fully written out. Section 12.3 verifies the normalized local estimates (12.31)–(12.33) on covering balls and then states "we can repeat the iteration argument on [54, p. 12]" and asserts (12.34) with a constant C0^N. Theorem 5.1 underpins Theorems 2.8, 2.9, 3.3 and 3.4, and the claimed linear-in-M form of the exponents in those theorems depends on the chain product being exactly e^{C0 N} = e^{C0 D(M + r_Ω^{-1})}, not e^{C0 N^2} or e^{C0 N I}. Please expand this step: either reproduce the propagation inequality from [54, Theorem 2.1] with its explicit dependence on the number of balls, or give a direct induction from (12.31)–(12.33) in which each step multiplies the L^ε norm by a universal constant and the number of steps is bounded by the N in Proposition 12.2. This is a load-bearing point because any superlinear growth in N would destroy the optimized form of the main estimates.

minor comments (4)

[Remark 2.3] The symbol after "f" in the sentence "(2.18) fails for n ≥ 2 and f /greaterornotdbleql0" is a typesetting artifact; it should read "f \not\equiv 0" (or the intended relation).
[Proof of Proposition 4.3] In the displayed integral after (10.6), the factor "(1-x)" should be "(1-|x|)" to match the weight d(x) for the unit ball.
[Theorems 2.9 and 3.4] In the OCR text the expressions "d(x)1− n/q" appear without proper superscripts; the intended expressions are d(x)^{1-n/q} in (2.19) and Theorem 3.4(iii).
[Proposition 12.2] In the proof of part (ii), the inequality in the claim (12.19) would be easier to follow if the authors explicitly noted that r ≤ (120 k_Ω)^{-1/α} implies r/52 ≥ k_Ω(3r/2)^{1+α}, which is needed in the displayed chain of inequalities.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central estimates are new derivations; the load-bearing Harnack input is an adapted, refereed prior-result dependency, not a re-derivation of the paper's own outputs.

full rationale

The paper's headline results (Theorems 2.1-2.9 and 3.1-3.4) are derived from classical Schauder/de Giorgi-Nash estimates, an optimized boundary weak Harnack inequality (Theorem 5.1), and explicit duality, scaling, and covering arguments. The optimized Harnack theorem is adapted from the authors' previous refereed papers [50] and [54]: Section 12.3 proves the local normalized estimates (12.31)-(12.33) and then states 'we can repeat the iteration argument on [54, p. 12]'. This is a legitimate dependency on published theorems with stated assumptions that do not include the present conclusions; [50, Theorem 1.1] and [54, Theorem 2.1] are parameter-free a priori results, so the self-citation is real evidence rather than a circular load-bearing premise. The chain-length bound Proposition 12.2 is proved in the paper and is what turns the local estimates into the global constant e^{C0(r_Omega^{-1}+M)D}; no equation in Theorem 5.1 is assumed as an input of itself. The optimality claims are established by explicit constructions, e.g. Propositions 4.1-4.4 and 11.1-11.2, in which M is the actual coefficient norm from (1.7); those examples are not fitted to the desired inequality. The proof of Theorem 5.1 is presented as a sketch, and the precise iteration constant from [54] is not written out, so a reader may want a fuller verification of that imported step; this is a rigor/exposition concern, not a circularity. Therefore no circular step is identified, and the score is 0.

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

The paper introduces no fitted parameters; its constants depend only on structural data (n, λ, Λ, q, α, ᾱ). The quantity M is a coefficient norm defined implicitly through the radius r0 in (3.2), with well-posedness proved in Proposition 5.3. The load-bearing imported premises are the optimized Harnack inequalities from the authors' prior works [50] and [54] (Theorem 5.1 and Theorem 12.3), the geometric chain lemma (Proposition 12.2), first-eigenvalue theory for divergence operators with L^q coefficients [9], [10], and the classical regularity and Hopf-type tools from [22] and [44]. The uniform eigenvalue positivity (2.7)/(3.13) is an honest structural assumption whose necessity the paper demonstrates. No new entities, forces, or conserved quantities are introduced; the 'multiplier' of earlier Landis proofs is explicitly avoided.

assumptions (6)

domain assumption Boundary weak Harnack inequality with constants depending only on normalized coefficient norms [50, Theorem 1.1] (stated as Theorem 12.3 here).
Imported from the authors' prior work; the adaptation to general C^{1,ᾱ} domains and to the r0-rescaling is given as a proof sketch in Section 12.3. The optimized estimates in Theorems 2.1, 2.8, 2.9, 3.3 and 3.4 rest on it.
standard math Harnack chain geometry: Proposition 12.2 constructs chains of balls with length N ≤ 2 + 6D/r and I ≤ c0(1 + D0/r)^n with controlled overlaps.
Proof provided in Section 12.3; uses only C^{1,ᾱ} flattening and covering arguments. Needed to propagate Harnack estimates across general domains.
standard math First eigenvalue theory for divergence operators with L^q coefficients [9], [10], including the characterization (12.1) and domain monotonicity.
Used in Theorems 2.3, 2.4, 2.6 and Proposition 2.5; properties listed in Section 12.1, cited from Chicco.
ad hoc to paper Uniform positivity of the first eigenvalue on a larger domain: λ1(-L, Ω̂) ≥ 0 in (2.7) and (3.13).
This is the key structural hypothesis; the paper shows it is necessary, since the solution of -Δu - λu = 1 in B1 explodes as λ ↑ λ1(-Δ, B1) (Section 2.1), and Proposition 2.7 shows failure of the Landis estimate when λ1 < 0.
standard math Regularity inheritance: weak solutions of (1.2) are C^{1,α} up to the boundary under (1.3)-(1.4), from [22, Sections 8.11, 8.29, 8.33] and [44, Section 5.5].
Used in the integration-by-parts argument (Lemma 12.4) and in the proofs of Theorems 2.2, 2.6, 2.8, 3.2, and 3.4.
standard math Classical C^{1,ν} and W^{1,q} interior and boundary estimates, Morrey embedding, and the doubling lemma [49, Lemma 5.1].
Used in the doubling-rescaling contradiction proof of Theorem 2.9 (Section 9) and in the inner steps of Lemma 8.1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Elliptic regularity estimates with optimized constants and applications." pith.science (2026). https://pith.science/paper/YGBFWLCK

@misc{pith2026241119367,
  author       = {Pith},
  title        = {Pith review of: Elliptic regularity estimates with optimized constants and applications},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YGBFWLCK}},
  note         = {Machine review of arXiv:2411.19367}
}

abstract

We revisit the classical theory of linear second-order uniformly elliptic equations in divergence form whose solutions have H\"older continuous gradients, and prove versions of the generalized maximum principle, the $C^{1,\alpha}$-estimate, the Hopf-Oleinik lemma, the boundary weak Harnack inequality and the differential Harnack inequality, in which the constant is optimized with respect to the norms of the coefficients of the operator and the size of the domain. Our estimates are complemented by counterexamples which show their optimality. We also give applications to the Landis conjecture and spectral estimates.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 53 canonical work pages

[54]

Sirakov, Ph

B. Sirakov, Ph. Souplet, The V´ azquez maximum principl e and the Landis conjecture for elliptic PDE with unbounded coeﬃcients, Adv. Math. 387 (2021), Paper No. 107838, 27 pp

work page 2021
[1]

Alvarado, D

R. Alvarado, D. Brigham, V. Maz’ya, M. Mitrea, E. Ziad´ e, On the regularity of domains satisfying a uniform hour-glass condition and a sharp version of the Hopf -Oleinik boundary point principle, Problems in mathematical analysis. No. 57. J. Math. Sci. (N.Y.) 176 (2011), 281-360

work page 2011
[2]

Apushkinskaya, A.I

D.E. Apushkinskaya, A.I. Nazarov, The normal derivativ e lemma and surrounding issues, Russian Math. Surveys, 77(2), (2022), 189-249

work page 2022
[3]

Arapostathis, A

A. Arapostathis, A. Biswas, D. Ganguly, Certain Liouvil le properties of eigenfunctions of elliptic oper- ators, Trans. Amer. Math. Soc. (6) 371 (2019), 4377-4409

work page 2019
[4]

Le Balc’h, Exponential bounds for gradient of solutio ns to linear elliptic and parabolic equations, J

K. Le Balc’h, Exponential bounds for gradient of solutio ns to linear elliptic and parabolic equations, J. Funct. Anal. 281(2021), Paper No. 109094, 29 pp

work page 2021
[5]

Le Balc’h, D

K. Le Balc’h, D. Souza, Quantitative unique continuatio n for real-valued solutions to second order elliptic equations in the plane, Preprint, arXiv: 2401.004 41

work page
[6]

Berestycki, L

H. Berestycki, L. Nirenberg, S.R.S. Varadhan, The princ ipal eigenvalue and maximum principle for second-order elliptic operators in general domains, Comm. Pure Appl. Math. 47 (1994), 47-92

work page 1994
[7]

Bourgain, C

J. Bourgain, C. Kenig, On localization in the continuous Anderson-Bernoulli model in higher dimension, Invent. Math. 161 (2005), 389-426

work page 2005

Show all 56 references

[8]

Brezis, X

H. Brezis, X. Cabr´ e, Some simple nonlinear PDE’s withou t solutions, Boll. Unione Mat. Ital. (8) 1-B (1999), 223–262

1999
[9]

Chicco, Principio di massimo generalizzato e valutaz ione del primo autovalore per problemi ellittici del secondo ordine di tipo variazionale, Ann

M. Chicco, Principio di massimo generalizzato e valutaz ione del primo autovalore per problemi ellittici del secondo ordine di tipo variazionale, Ann. Mat. Pura Appl. (4) 87 (1970), 1-9

1970
[10]

Chicco, Some properties of the ﬁrst eigenvalue and th e ﬁrst eigenfunction of linear second order elliptic partial diﬀerential equations in divergence form , Boll

M. Chicco, Some properties of the ﬁrst eigenvalue and th e ﬁrst eigenfunction of linear second order elliptic partial diﬀerential equations in divergence form , Boll. Un. Mat. Ital. (4) 5 (1972), 245-254

1972
[11]

U. Das, Y. Pinchover, The Landis conjecture via Liouvil le comparison principle and criticality theory, Preprint, arXiv: 2405.11695

arXiv
[12]

Davey, Quantitative unique continuation for Schr¨ o dinger operators, J

B. Davey, Quantitative unique continuation for Schr¨ o dinger operators, J. Funct. Anal. 279 (2020), 108566,

2020
[13]

Davey, C

B. Davey, C. Kenig, J.-N. Wang, The Landis conjecture fo r variable coeﬃcient second-order elliptic PDEs, Trans. Amer. Math. Soc. 369 (2017), 8209-8237

2017
[14]

Davey, J.-N

B. Davey, J.-N. Wang, Landis’ conjecture for general se cond order elliptic equations with singular lower order terms in the plane, J. Diﬀerential Equations 268 (2020), 977-1042

2020
[15]

Davey, J

B. Davey, J. Zhu, Quantitative uniqueness of solutions to second-order elliptic equations with singular lower order terms, Comm. Partial Diﬀerential Equations 44 (2019), 1217-1251

2019
[16]

Drabek, J

P. Drabek, J. Milota, Methods of nonlinear analysis, Bi rkh¨ auser Adv. Texts Basler, Basel, 2013, x+649 pp

2013
[17]

Du, Order structure and topological methods in nonli near partial diﬀerential equations, Series in Partial Diﬀerential Equations and Applications, World Sci entiﬁc Publishing, 2006

Y. Du, Order structure and topological methods in nonli near partial diﬀerential equations, Series in Partial Diﬀerential Equations and Applications, World Sci entiﬁc Publishing, 2006. x+190 pp

2006
[18]

El Berdan, J

N. El Berdan, J. D ´ ıaz, J.-M. Rakotoson, The uniform Hop f inequality for discontinuous coeﬃcients and optimal regularity in BMO for singular problems, J. Math. Anal. Appl. 437 (2016) 350-379

2016
[19]

Escauriaza, C

L. Escauriaza, C. Kenig, G. Ponce, L. Vega, On uniquenes s properties of solutions of Schr¨ odinger equations, Comm. Partial Diﬀerential Equations 31 (2006), 1811-1823

2006
[20]

Felmer, A

P. Felmer, A. Quaas, B. Sirakov, Solvability of nonline ar elliptic equations with gradient terms, J. Diﬀerential Equations 254 (2013), 4327-4346

2013
[21]

Fonte Sanchez, P

C. Fonte Sanchez, P. Gabriel, S. Mischler, On the Krein- Rutman theorem and beyond, Preprint, arXiv: 2305.06652
[22]

Gilbarg, N

D. Gilbarg, N. Trudinger, Elliptic Partial Diﬀerentia l Equations of Second Order, 2nd edition, Springer Verlag, 2001

2001
[23]

Ginibre, G

J. Ginibre, G. Velo, The Cauchy problem in local spaces f or the complex Ginzburg Landau equation. II. Contraction methods, Comm. Math. Phys. 187 (1997), 45-79

1997
[24]

Haque, T

M. Haque, T. Ogawa, R. Sato, Existence of weak solutions to a convection-diﬀusion equation in a uniformly local Lebesgue space, Commun. Pure Appl. Anal. 19 (2020), 677-697

2020
[25]

Ignatova, I

M. Ignatova, I. Kukavica, L. Ryzhik, The Harnack inequa lity for second-order elliptic equations with divergence-free drifts, Commun. Math. Sci. 12 (2014), 681-694

2014
[26]

Ishige, R

K. Ishige, R. Sato, Heat equation with a nonlinear bound ary condition and uniformly local Lr spaces, Discrete Contin. Dyn. Syst. 36, (2016) 2627-2652. ELLIPTIC REGULARITY ESTIMATES WITH OPTIMIZED CONSTANTS 53

2016
[27]

Julin, Generalized Harnack inequality for semiline ar elliptic equations, J

V. Julin, Generalized Harnack inequality for semiline ar elliptic equations, J. Math. Pures Appl. (9) 106 (2016), 877-904

2016
[28]

Kato, The Cauchy problem for quasi-linear symmetric hyperbolic systems, Arch

T. Kato, The Cauchy problem for quasi-linear symmetric hyperbolic systems, Arch. Rational Mech. Anal. 58 (1975), 181-205

1975
[29]

Kenig, Some recent applications of unique continuat ion

C. Kenig, Some recent applications of unique continuat ion. Recent developments in nonlinear partial diﬀerential equations, 25-56, Contemp. Math., 439, Amer. M ath. Soc., Providence, RI, 2007

2007
[30]

Kenig, L

C. Kenig, L. Silvestre, J.-N. Wang, On Landis’ conjectu re in the plane, Comm. Partial Diﬀerential Equations 40 (2015), 766-789

2015
[31]

Kenig, J.-N

C. Kenig, J.-N. Wang, Quantitative uniqueness estimat es for second order elliptic equations with un- bounded drift, Math. Res. Lett. 22 (2015), 1159-1175

2015
[32]

Klimsiak, T

T. Klimsiak, T. Komorowski, Hopf type lemmas for subsol utions of integro-diﬀerential equations, Bernoulli 29 (2023), 1435-1463

2023
[33]

Kondratiev, E

V. Kondratiev, E. Landis, On qualitative properties of solutions of a nonlinear equation of second order (Russian), Mat. Sb. (N.S.) 135 (177) (1988), 346-360; translation in Math. USSR-Sb. 63 (1989), 337-350

1988
[34]

Konkov, On solutions of quasilinear elliptic ineq ualities containing terms with lower-order deriva- tives, Nonlinear Anal

A.A. Konkov, On solutions of quasilinear elliptic ineq ualities containing terms with lower-order deriva- tives, Nonlinear Anal. 90 (2013), 121-134

2013
[35]

Krylov, Nonlinear elliptic and parabolic equations of second order, Reidel, Dordrecht, 1987

N. Krylov, Nonlinear elliptic and parabolic equations of second order, Reidel, Dordrecht, 1987

1987
[36]

Ladyzhenskaya, N.N

O.A. Ladyzhenskaya, N.N. Ural’tseva, Linear and quasi linear elliptic equations. Academic Press, New York-London, 1968

1968
[37]

Lieberman, Regularized distance and its applicatio ns, Paciﬁc J

G. Lieberman, Regularized distance and its applicatio ns, Paciﬁc J. Math. 117 (1985), 329-352

1985
[38]

Li, Gradient estimates and Harnack inequalities for nonlinear parabolic and nonlinear elliptic equa- tions on Riemannian manifolds, J

J. Li, Gradient estimates and Harnack inequalities for nonlinear parabolic and nonlinear elliptic equa- tions on Riemannian manifolds, J. Funct. Anal. 100 (1991) 233-256

1991
[39]

Li, S.-T

P. Li, S.-T. Yau, On the parabolic kernel of the Schr¨ odi nger operator, Acta Math. 156 (1986) 153-201

1986
[40]

Y. Li, L. Nirenberg, Generalization of a well known ineq uality, in: Progress in Nonlinear Diﬀerential Equations and Their Applications, Vol. 66, 365-370, 2005

2005
[41]

Logunov, E

A. Logunov, E. Malinnikova, N. Nadirashvili, F. Nazaro v, The Landis conjecture on exponential decay, Preprint, arXiv: 2007.07034

2007 arXiv
[42]

B. Ma, F. Zeng, Hamilton-Souplet-Zhang’s gradient est imates and Liouville theorems for a nonlinear parabolic equation, C. R. Math. Acad. Sci. Paris 356 (2018), 550-557

2018
[43]

Maekawa, Y

Y. Maekawa, Y. Terasawa, The Navier-Stokes equations w ith initial data in uniformly local Lp spaces Diﬀer. Integral Equ. 19 (2006), 369-400

2006
[44]

Morrey, Multiple integrals in the calculus of vari ations, Classics Math

C.B. Morrey, Multiple integrals in the calculus of vari ations, Classics Math. Springer-Verlag, Berlin, 2008

2008
[45]

M¨ uller, Diﬀerential Harnack inequalities and the R icci ﬂow

R. M¨ uller, Diﬀerential Harnack inequalities and the R icci ﬂow. EMS Series of Lectures in Mathematics. Z¨ urich, 2006

2006
[46]

Nornberg, C 1,α regularity for fully nonlinear elliptic equations with sup erlinear growth in the gradi- ent, J

G. Nornberg, C 1,α regularity for fully nonlinear elliptic equations with sup erlinear growth in the gradi- ent, J. Math. Pures Appl. (9) 128 (2019), 297-329

2019
[47]

Nornberg, B

G. Nornberg, B. Sirakov, A priori bounds and multiplici ty for fully nonlinear equations with quadratic growth in the gradient, J. Funct. Anal. 276 (2019), 1806-1852

2019
[48]

Orsina, A.C

L. Orsina, A.C. Ponce, Hopf potentials for the Schr¨ odi nger operator, Anal. PDE 11 (2018), 2015-2047

2018
[49]

Pol´ aˇ cik, P

P. Pol´ aˇ cik, P. Quittner, Ph. Souplet, Singularity an d decay estimates in superlinear problems via Liouville-type theorems. I. Elliptic equations and system s, Duke Math. J. 139 (2007), 555-579

2007
[50]

Rend´ on, B

F. Rend´ on, B. Sirakov, M. Soares, Boundary weak Harnac k estimates and regularity for elliptic PDE in divergence form, Nonlinear Anal. 235 (2023), Paper No. 113331

2023
[51]

Rossi, The Landis conjecture with sharp rate of decay , Indiana Univ

L. Rossi, The Landis conjecture with sharp rate of decay , Indiana Univ. Math. J. 70 (2021), 301-324

2021
[52]

Sirakov, Boundary Harnack estimates and quantitati ve strong maximum principles for uniformly elliptic PDE, Int

B. Sirakov, Boundary Harnack estimates and quantitati ve strong maximum principles for uniformly elliptic PDE, Int. Math. Res. Notices (2018) (24) 7457-7482

2018
[53]

Sirakov, Global integrability and weak Harnack esti mates for elliptic PDE in divergence form, Anal- ysis and PDE 15 (2022), 2849-2868

B. Sirakov, Global integrability and weak Harnack esti mates for elliptic PDE in divergence form, Anal- ysis and PDE 15 (2022), 2849-2868

2022
[55]

Trudinger, Linear elliptic equations with measurab le coeﬃcients, Ann

N. Trudinger, Linear elliptic equations with measurab le coeﬃcients, Ann. Scuola Norm. Sup. Pisa (3) 27, 265-308 (1973)

1973
[56]

Zhao, Green function for Schr¨ odinger operator an d conditioned Feynman-Kac gauge, J

Z.X. Zhao, Green function for Schr¨ odinger operator an d conditioned Feynman-Kac gauge, J. Math. Anal. Appl. 116 (1986), 309-334. 54 SIRAKOV AND SOUPLET PUC-Rio, Departamento de Matematica, Rua Marqu ˆes de S ˜ao Vicente 225, G ´avea, Rio de Janeiro - CEP 22451-900, Brazil Em...

1986

Pith tools

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

[1] [54]

Sirakov, Ph

B. Sirakov, Ph. Souplet, The V´ azquez maximum principl e and the Landis conjecture for elliptic PDE with unbounded coeﬃcients, Adv. Math. 387 (2021), Paper No. 107838, 27 pp

work page 2021

[2] [1]

Alvarado, D

R. Alvarado, D. Brigham, V. Maz’ya, M. Mitrea, E. Ziad´ e, On the regularity of domains satisfying a uniform hour-glass condition and a sharp version of the Hopf -Oleinik boundary point principle, Problems in mathematical analysis. No. 57. J. Math. Sci. (N.Y.) 176 (2011), 281-360

work page 2011

[3] [2]

Apushkinskaya, A.I

D.E. Apushkinskaya, A.I. Nazarov, The normal derivativ e lemma and surrounding issues, Russian Math. Surveys, 77(2), (2022), 189-249

work page 2022

[4] [3]

Arapostathis, A

A. Arapostathis, A. Biswas, D. Ganguly, Certain Liouvil le properties of eigenfunctions of elliptic oper- ators, Trans. Amer. Math. Soc. (6) 371 (2019), 4377-4409

work page 2019

[5] [4]

Le Balc’h, Exponential bounds for gradient of solutio ns to linear elliptic and parabolic equations, J

K. Le Balc’h, Exponential bounds for gradient of solutio ns to linear elliptic and parabolic equations, J. Funct. Anal. 281(2021), Paper No. 109094, 29 pp

work page 2021

[6] [5]

Le Balc’h, D

K. Le Balc’h, D. Souza, Quantitative unique continuatio n for real-valued solutions to second order elliptic equations in the plane, Preprint, arXiv: 2401.004 41

work page

[7] [6]

Berestycki, L

H. Berestycki, L. Nirenberg, S.R.S. Varadhan, The princ ipal eigenvalue and maximum principle for second-order elliptic operators in general domains, Comm. Pure Appl. Math. 47 (1994), 47-92

work page 1994

[8] [7]

Bourgain, C

J. Bourgain, C. Kenig, On localization in the continuous Anderson-Bernoulli model in higher dimension, Invent. Math. 161 (2005), 389-426

work page 2005