Strichartz Estimates for the Schr\"odinger Equation with a Measure-Valued Potential

M. Burak Erdogan; Michael Goldberg; William R. Green

REVIEW 2 major objections 5 minor 21 references

Strichartz Estimates for the Schr\"odinger Equation with a Measure-Valued Potential

T0 review · 2 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 The paper proves that compactly supported measure potentials of dimension > n − (1 + 1/(n−1)) preserve Schrödinger Strichartz estimates, with the possible exception of the endpoint pair, provided the Hamiltonian has no zero resonance and…

desk verdict New Strichartz estimates for measure-valued potentials in n≥4, with a clean high-energy method that rests on one under-proved restriction-theoretic transfer. read the letter →

arxiv 1908.02903 v1 pith:TIR3WTLD submitted 2019-08-08 math.AP

M. Burak Erdogan , Michael Goldberg , William R. Green This is my paper · ORCID

classification math.AP MSC 35Q4135P2542B37

keywords Strichartzestimatesmeasure-valuedpotentialsSchrödingerequationlocaldecayresolventFourierrestrictionsingularfractalmeasures

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper establishes that the full family of Strichartz inequalities for the Schrödinger equation survives when the free Hamiltonian $-\Delta$ is perturbed by a compactly supported signed measure $\mu$, provided the measure has dimension $\alpha > n-(1+\frac{1}{n-1})$ and the perturbed operator has no eigenvalues at $\lambda\ge 0$ and no resonance at zero. This covers potentials that are not functions at all, including the surface measure of a compact hypersurface in $\mathbb{R}^n$, a regime where $L^1\to L^\infty$ dispersive estimates can fail even though Strichartz estimates still hold. The intermediate step is a local decay estimate in $L^2(\mu)$: both the free evolution and the perturbed evolution, projected onto the continuous spectrum, have square-integrable-in-time $L^2(\mu)$ norm. If the theorem is correct, the standard Strichartz toolbox applies to these rough potentials with only the endpoint pair $(2, 2n/(n-2))$ possibly excluded.

What carries the argument

The load-bearing object is the family of free limiting resolvents $R_0^\pm(\lambda^2)\mu$ acting on $L^2(\mu)$. Uniform bounds on the difference $R_0^+(\lambda^2)\mu - R_0^-(\lambda^2)\mu$, and on the analogous difference for the perturbed resolvent, are equivalent to the local decay estimates by a standard spectral-theoretic argument; the Strichartz estimates then follow by writing the perturbed evolution in variation-of-parameters form and using the dual local decay bound together with the free Strichartz estimates. At low energy, the bounds come from compactness of $R_0^+(\lambda^2)\mu$ on $L^2(\mu)$, obtained through the compact embedding of $\dot H^1(\mathbb{R}^n)$ into $L^2(\mu)$. At high energy, the free resolvent is split into a sphere-restriction term, controlled by a sharp $L^2$ Fourier-restriction estimate scaled as $R^{\alpha/(2n)}$, and a principal-value integral over the sphere radius, controlled by a single integration by parts; the necessary derivative bound for the spherical restriction follows from the compact support of $\mu$.

What would settle it

A concrete test is the spherical restriction bound used in Section 4: for $\mu_R = R^\alpha\mu(\cdot/R)$ and $g\in L^2(S^{n-1})$, the estimate $\|\widehat{g}\|_{L^2(\mu_R)} \lesssim R^{\alpha/(2n)}\|g\|_2$ must hold with no extra power of $R$; finding any $g$ with growth $R^{\alpha/(2n)+\delta}$, $\delta>0$, would make the exponent in (22) nonnegative and invalidate the proof. Independently, a compactly supported measure satisfying the spectral assumptions for which the local decay bound (5) fails would refute the theorem.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for $n\ge 3$, if $\mu$ is a compactly supported signed measure on $\mathbb{R}^n$ of dimension $\alpha > n-(1+\frac{1}{n-1})$, and if $-\Delta+\mu$ has no resonance at zero and no eigenvalues at any $\lambda\ge 0$, then for every $f\in L^2(\mathbb{R}^n)$, $$ \|e^{it\$\Delta$}f\|_{$L^{2}$_t $L^{2}$(\mu)} \lesssim \|f\|$_2^{2}$, \qquad \|e^{it(-\$\Delta$+\mu)}P_{ac}f\|_{$L^{2}$_t $L^{2}$(\mu)} \lesssim \|f\|$_2^{2}$, $$ and the Strichartz inequalities $\|e^{it(-\Delta+\mu)}P_{ac}f\|_{L^p_t L^q_x}\lesssim\|f\|_2$ hold for all admissible pairs with $2/p+n/q=n/2$ and $p>2$. The possible exception is the endpoint $(2, 2n/(n-2))$. The dimensional threshold is exactly what makes the high-energy resolvent exponent $n-2-\alpha(n-1)/n$ negative, which is how the surface-measure case $\alpha=n-1$ becomes admissible in every dimension $n\ge 3$.

Load-bearing premise

The load-bearing premise is that the sharp $L^2$ Fourier-restriction estimate for the paraboloid transfers to the unit sphere with exactly the same $R^{\alpha/(2n)}$ scaling; the paper states that this transfer is well known, but it gives no proof, and any additional power of $R$ would make the high-energy resolvent exponent $n-2-\alpha(n-1)/n$ nonnegative and break the argument.

Editorial extensions

If this is right

In every dimension $n\ge 3$, the surface measure of a compact hypersurface is an admissible potential, so Strichartz estimates with $p>2$ hold once the spectral assumptions are satisfied.
The local decay bounds hold in the strong form $\|e^{itH}P_{ac}f\|_{L^2_t L^2(\mu)} \lesssim \|f\|_2^2$ for both $H=-\Delta$ and $H=-\Delta+\mu$.
The only admissible Strichartz pair that can be lost is the endpoint $(2, 2n/(n-2))$; all others are preserved.
The theorem extends the previously understood measure-potential dispersive theory from three dimensions to all $n\ge 4$.

Reading between the lines

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

The proof's mechanism is scale-sensitive: the dimensional threshold is set by the $R^{\alpha/(2n)}$ scaling in the restriction estimate, so any future improvement in that scaling would automatically lower the threshold, and a counterexample with a larger power would break this particular argument.
The same high-energy decomposition should work for other uniformly convex smooth surfaces in place of the sphere, since only the restriction estimate and the dimension-based integral bounds are used.
The endpoint Strichartz pair is left open, not ruled out; deciding whether a surface-measure potential can actually destroy the endpoint would require a separate argument or a counterexample.

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.

Desk Editor's Note

read the letter

Two-line take: this paper proves Strichartz estimates for Schrödinger operators with α-dimensional measure-valued potentials in R^n, n≥3, under a spectral assumption. The n≥4 part is new, includes the hypersurface case α=n−1, and the high-energy argument via Du–Zhang's restriction estimate is a genuinely different tool that avoids the L^1→L^∞ dispersive estimates known to fail without smoothness. What's good: the self-adjointness and low-energy sections are careful. The compact embedding of H^1 into L^2(μ) for α>n−2 is proved cleanly; the KLMN form bound is handled properly. The low-energy resolvent invertibility uses a nice trick with mollification to rule out positive eigenvalues. The high-energy observation that μf is not a generic H^{−1} element, so the free resolvent actually decays on L^2(μ), is the heart of the paper and is rewarding. Now the soft spot. Section 4, Theorem 4.1, is load-bearing and under-proved. The bound (20) follows from Du–Zhang's paraboloid estimate by replacing the paraboloid with the sphere, stated as 'well known'. For Lebesgue measures on R^n that equivalence is standard, but here the target is the rescaled fractal measure μ_R, and a nonlinear change of variables on the frequency side does not obviously preserve the L^2(μ_R) norm of the extension. If the sphere estimate carried any R^ε loss, the exponent in (22) would become n−2 − α(n−1)/n + ε, which at the threshold α = n − (1 + 1/(n−1)) equals −1 + 1/n + ε; any ε > 1 − 1/n kills the decay. The derivative bound (21) is also only sketched. This is not obviously fatal: the transfer is plausible and likely repairable (Luca–Rogers may already contain it), and the rest of the proof would then go through. But as written it is a gap in the main estimate. Citation pattern is fine; self-citations are for the R^3 base case and a Helmholtz theorem. No circularity. Bottom line: this is a serious paper that deserves a serious referee. A referee should ask for a precise reference or proof of the sphere restriction estimate and a fuller derivation of (21). If those are supplied—or the theorem restated with a slightly smaller α range—the result stands. I'd send it to peer review without hesitation, and I'd cite it if I worked on singular potentials.

Referee Report

2 major / 5 minor

Summary. The paper proves local decay and Strichartz estimates for the Schrödinger evolution generated by H = -Δ + μ in R^n, n ≥ 3, where μ is a compactly supported signed measure of dimension α > n - (1 + 1/(n-1)). Under the spectral assumption that H has no eigenvalues at nonnegative energies and no resonance at zero, Theorem 1.1 establishes the local decay bounds (4)-(5) and the Strichartz inequalities (6) for all admissible pairs with p > 2, with the endpoint (2, 2n/(n-2)) possibly excluded. The proof follows the Rodnianski-Schlag framework: self-adjointness and compactness of the inclusion of ẒH^1 into L^2(μ) are obtained via the KLMN theorem and a translation estimate (Section 2); low-energy resolvent bounds are proved by Fredholm theory and a bootstrapping argument (Section 3); high-energy decay of the free resolvent on L^2(μ) is derived from a fractal Fourier restriction estimate attributed to Du and Zhang, with a paraboloid-to-sphere transfer asserted as well known (Section 4).

Significance. If the high-energy restriction step is fully justified, this is a substantial contribution: it extends Strichartz estimates to singular measure-valued potentials in all dimensions n ≥ 3, including hypersurface-supported potentials in higher dimensions where L^1 → L^∞ dispersive estimates are known to fail. The self-adjointness argument via the KLMN theorem and the compact embedding of ẒH^1 into L^2(μ) are carefully presented, and the reduction of local decay to uniform resolvent bounds is standard and clearly explained. The main caveat is that the paper's central new input, the high-energy decay of the free resolvent, depends on an unproved transfer of the Du-Zhang estimate from the paraboloid to the sphere with the identical fractal-measure scaling; this is load-bearing and currently leaves Theorem 1.1 conditional. The manuscript otherwise appears coherent and the use of recent sharp restriction estimates is appropriate and innovative.

major comments (2)

[Section 4, Theorem 4.1, display (20)] The assertion that the Du-Zhang L^2 restriction estimate for the paraboloid transfers to the unit sphere with the identical power R^{α/(2n)} is stated as 'well known' but is not automatic for the fractal measure μ_R. A nonlinear change of variables from a sphere cap to the paraboloid does not preserve L^2(μ_R), and the standard equivalence of restriction estimates for smooth surfaces is usually proved for Lebesgue norms. This is load-bearing: the exponent in (22) is n - 2 - α(n-1)/n, and the hypothesis α > n - (1 + 1/(n-1)) only makes this a small negative number when α is close to the threshold. A loss of even R^ε in (20) would make the exponent positive for α sufficiently close to the threshold and would destroy the high-energy decay on which Theorem 1.1 depends. Please provide a proof or a precise citation for the sphere version with identical scaling, or revise the argument accordingly.
[Section 4, inequality (21)] The bound on the outward normal gradient of \widehat{μf} is justified only by a brief comment that compact support makes the L^2(μ) norm of (1+|x|)f comparable to that of f. As written this is not immediate, because differentiating \widehat{μf} with respect to ξ produces the measure x_j μ, and one must verify that each x_j μ satisfies the α-dimensional bound with a constant uniform in j. The estimate is plausibly repairable by applying (20) to each x_j μ, but the details should be written out because (21) is used in the integration by parts controlling the principal-value part (23).

minor comments (5)

[References] Reference [5] lists the year as '2109'; this should be '2019'.
[Section 4, surface measure term] The phrase 'the T*T composition of the operator in (20)' should define the operator T explicitly before using T*T notation.
[Section 4, display (20)] The notation L^2(RS^{n-1}) should be defined as the L^2 space with respect to the induced surface measure on the sphere of radius R.
[Lemma 3.1] Lemma 3.1 assumes a 'real-valued measure' while Theorem 1.1 states 'signed measure'; if signed measures are understood to be real-valued by convention, this should be stated explicitly.
[Introduction] The text says the results for R^n, n ≥ 4 are new, but Theorem 1.1 is stated for all n ≥ 3; the status of the n = 3 case relative to [10] and [15] should be clarified.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Theorem 1.1 follows from external restriction/resolvent frameworks; self-citations are not load-bearing.

full rationale

The main claim (Theorem 1.1) is obtained from uniform resolvent bounds (Theorem 1.2), which are assembled from low-energy (Lemma 3.1) and high-energy (Theorem 4.1) estimates. The high-energy estimate (22) is derived from the Du–Zhang paraboloid restriction theorem with a stated transfer to the sphere; even if that transfer is under-proved, it is an external input rather than a re-statement of the paper's conclusion. Local decay is derived via Kato's argument and Strichartz via Rodnianski–Schlag, both external. Self-citations ([10], [11]) occur only for the R^3 base case and a supporting Helmholtz-theorem step; neither is used as a uniqueness theorem or as a substitute for the main derivation, and the power α/(2n) in (20) comes from Du–Zhang, not from the authors' previous work. No fitted parameter is renamed as a prediction and no quantity is defined in terms of the quantity it is used to prove. The unproved 'well known' paraboloid-to-sphere transfer and the sketched derivative bound (21) are potential correctness gaps, but they do not make the argument circular.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The central claim rests on standard functional-analytic theorems (KLMN, Fredholm alternative, Christ-Kiselev), two deep external restriction results (Du-Zhang, and a Helmholtz theorem of Goldberg), and the explicitly stated spectral assumption. No free parameters are fitted to data; the dimension α and constants C_μ, M are hypotheses. The only ad hoc element is the unproved paraboloid-to-sphere transfer, which is flagged.

assumptions (7)

standard math KLMN theorem for construction of the self-adjoint form sum -Δ+μ
Used in Proposition 2.1 after proving the form bound (16) with a<1.
standard math Fredholm alternative for compact operators
Used in Lemma 3.1 and Corollary 2.6 to characterize invertibility of I+R0+(λ^2)μ.
standard math Christ-Kiselev lemma for time-ordered integrals
Used in the final step of Theorem 1.1 to restrict the integral to 0≤s≤t; requires p>2.
domain assumption Du-Zhang sharp L2 restriction estimate for the paraboloid (Theorem 2.3 of [5])
External deep theorem; the entire high-energy argument in Theorem 4.1 is built on it.
domain assumption Goldberg's Helmholtz theorem with L^p data (Theorem 2 of [11])
Used in Lemma 3.1 to bootstrap null-space functions to L^2(R^n).
domain assumption Spectral hypothesis: no resonance at zero and no eigenvalues at λ≥0
Stated as a hypothesis of Theorem 1.1; needed for the uniform low-energy resolvent bound in Lemma 3.1.
ad hoc to paper Paraboloid-to-sphere transfer of the Du-Zhang estimate with identical R^{α/(2n)} scaling
Asserted as 'well known' in Section 4 without proof or citation; load-bearing for the negative exponent in (22).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Strichartz Estimates for the Schr\"odinger Equation with a Measure-Valued Potential." pith.science (2026). https://pith.science/paper/TIR3WTLD

@misc{pith2026190802903,
  author       = {Pith},
  title        = {Pith review of: Strichartz Estimates for the Schr\"odinger Equation with a Measure-Valued Potential},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TIR3WTLD}},
  note         = {Machine review of arXiv:1908.02903}
}

abstract

We prove Strichartz estimates for the Schr\"odinger equation in $\mathbb R^n$, $n\geq 3$, with a Hamiltonian $H = -\Delta + \mu$. The perturbation $\mu$ is a compactly supported measure in $\mathbb R^n$ with dimension $\alpha > n-(1+\frac{1}{n-1})$. The main intermediate step is a local decay estimate in $L^2(\mu)$ for both the free and perturbed Schr\"odinger evolution.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[1]

Agmon and L

S. Agmon and L. H¨ ormander, Asymptotic properties of solutions of diﬀerential equatio ns with simple characteristics. J. Anal. Math. 30 (1976), no. 1, 1–38

work page 1976
[2]

Cornean, A

H. Cornean, A. Michelangeli and K. Yajima, Two-dimensional Schr¨ odinger operators with point interactions: threshold expansions, zero modes and Lp-boundedness of wave operators. Rev. Math. Phys. 31 (2019), no. 4, 1950012

work page 2019
[3]

D’Ancona, V

P. D’Ancona, V. Pierfelice, and A. Teta, Dispersive estimate for the Schr¨ odinger equation with point interactions. Math. Meth. Appl. Sci. 29 (2006), no. 3, 209–323

work page 2006
[4]

Dell’Antonio, A

G. Dell’Antonio, A. Michelangeli, R. Scandone, and K. Yajima, Lp-boundedness of wave operators for the three-dimensional multi-centre point interaction . Ann. Henri Poincar´ e19 (2018), no. 1, 283–322

work page 2018
[5]

Du and R

X. Du and R. Zhang, Sharp L2 estimates of the Schr¨ odinger maximal function in higher di mensions. Ann. Math. 189 (2109), no. 3, 837–861

work page
[6]

M. B. Erdogan, M. Goldberg and W. Schlag Strichartz and smoothing estimates for Schr¨ odinger operators with large magnetic potentials in R3, J. Eur. Math. Soc. 10 (2008), no. 2, 507-531

work page 2008
[7]

M. B. Erdogan, W. Green, Dispersive estimate for the Schr¨ odinger equation for C n´ 3 2 potentials in odd dimensions. Intl. Math. Res. Not. 2010 (2010), no. 13, 2532–2565

work page 2010
[8]

Ginibre and G

J. Ginibre and G. Velo, The global Cauchy problem for the nonlinear Schr¨ odinger eq uation revisited. Ann. Inst. H. Poincar´ e Anal. Non Lin´ eaire2 (1985), no. 4, 309–327

work page 1985

Show all 21 references

[9]

Goldberg, Strichartz estimates for the Schr¨ odinger equation with ti me-periodic Ln{2 potentials, J

M. Goldberg, Strichartz estimates for the Schr¨ odinger equation with ti me-periodic Ln{2 potentials, J. Funct. Anal. 256 (2009), 718–746

2009
[10]

Goldberg, Dispersive estimates for Schr¨ odinger operators with meas ure-valued potentials in R3

M. Goldberg, Dispersive estimates for Schr¨ odinger operators with meas ure-valued potentials in R3. Indiana Univ. Math. J. 61 (2012), no. 6, 2123-2141

2012
[11]

Goldberg, The Helmholtz equation with Lp data and Bochner-Riesz multipliers

M. Goldberg, The Helmholtz equation with Lp data and Bochner-Riesz multipliers . Mat. Res. Lett. 23 (2016), no. 6, 1665-1679

2016
[12]

Goldberg and M

M. Goldberg and M. Visan A counterexample to dispersive estimates for Schr¨ odinger operators in higher dimensions , Comm. Math. Phys. 266 (2006), no. 1, 211–238

2006
[13]

Journ´ e, A

J.-L. Journ´ e, A. Soﬀer, and C. Sogge, Decay estimates for Schr¨ odinger operators. Comm. Pure Appl. Math. 44 (1991) no. 5, 573–604

1991
[14]

Kato, Wave operators and similarity for some non-selfadjoint ope rators

T. Kato, Wave operators and similarity for some non-selfadjoint ope rators. Math. Ann. 162 (1965/66), 258–279. 16 M. BURAK ERDO ˘GAN, MICHAEL GOLDBERG, WILLIAM R. GREEN

1965
[15]

Keel, and T

M. Keel, and T. Tao. Endpoint Strichartz estimates , Amer. J. Math. 120 (1998), no. 5, 955-980

1998
[16]

Koch and D

H. Koch and D. Tataru, Carleman estimates and absence of embedded eigenvalues . Comm. Math. Phys. 267 (2006), no. 2, 419–449

2006
[17]

Luc` a and K

R. Luc` a and K. Rogers, Average decay of the Fourier transform of measures with appl ications, J. Eur. Math. Soc. 21 (2019), no. 2, 465–506

2019
[18]

Reed and B

M. Reed and B. Simon, Methods of Modern Mathematical Physics . II. Fourier Analysis, Self Ad- jointness. Academic Press [Harcourt Brace Jovanovich, Publisher s], New York–London, 1975

1975
[19]

Rodnianski and W

I. Rodnianski and W. Schlag, Time decay for solutions of Schr¨ odinger equations with rou gh and time-dependent potentials. Invent. Math. 155 (2004), no. 3, 451–513

2004
[20]

Schechter, Spectra of Partial Diﬀerential Operators

M. Schechter, Spectra of Partial Diﬀerential Operators. No rth-Holland, Amsterdam, 1971

1971
[21]

Strichartz, Restrictions of Fourier transforms to quadratic surfaces a nd decay of solutions of wave equations, Duke Journal of Mathematics, 44 (1977), 705–714

R. Strichartz, Restrictions of Fourier transforms to quadratic surfaces a nd decay of solutions of wave equations, Duke Journal of Mathematics, 44 (1977), 705–714. Department of Mathematics, University of Illinois, Urbana , IL 61801, U.S.A. E-mail address : berdogan@illinois.e...

1977

Pith tools

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

[1] [1]

Agmon and L

S. Agmon and L. H¨ ormander, Asymptotic properties of solutions of diﬀerential equatio ns with simple characteristics. J. Anal. Math. 30 (1976), no. 1, 1–38

work page 1976

[2] [2]

Cornean, A

H. Cornean, A. Michelangeli and K. Yajima, Two-dimensional Schr¨ odinger operators with point interactions: threshold expansions, zero modes and Lp-boundedness of wave operators. Rev. Math. Phys. 31 (2019), no. 4, 1950012

work page 2019

[3] [3]

D’Ancona, V

P. D’Ancona, V. Pierfelice, and A. Teta, Dispersive estimate for the Schr¨ odinger equation with point interactions. Math. Meth. Appl. Sci. 29 (2006), no. 3, 209–323

work page 2006

[4] [4]

Dell’Antonio, A

G. Dell’Antonio, A. Michelangeli, R. Scandone, and K. Yajima, Lp-boundedness of wave operators for the three-dimensional multi-centre point interaction . Ann. Henri Poincar´ e19 (2018), no. 1, 283–322

work page 2018

[5] [5]

Du and R

X. Du and R. Zhang, Sharp L2 estimates of the Schr¨ odinger maximal function in higher di mensions. Ann. Math. 189 (2109), no. 3, 837–861

work page

[6] [6]

M. B. Erdogan, M. Goldberg and W. Schlag Strichartz and smoothing estimates for Schr¨ odinger operators with large magnetic potentials in R3, J. Eur. Math. Soc. 10 (2008), no. 2, 507-531

work page 2008

[7] [7]

M. B. Erdogan, W. Green, Dispersive estimate for the Schr¨ odinger equation for C n´ 3 2 potentials in odd dimensions. Intl. Math. Res. Not. 2010 (2010), no. 13, 2532–2565

work page 2010

[8] [8]

Ginibre and G

J. Ginibre and G. Velo, The global Cauchy problem for the nonlinear Schr¨ odinger eq uation revisited. Ann. Inst. H. Poincar´ e Anal. Non Lin´ eaire2 (1985), no. 4, 309–327

work page 1985