Pointwise decay for semilinear wave equations in $\mathbb{R}^{!+3}$

Shiwu Yang

REVIEW 3 major objections 4 minor 26 references

Pointwise decay for semilinear wave equations in $\mathbb{R}^{!+3}$

T0 review · 3 major / 4 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 For powers above (1+√17)/2, the pointwise decay of the defocusing semilinear wave equation matches linear waves, and energy scattering follows for p>2.3542.

desk verdict Serious progress on pointwise decay and scattering below the Strauss exponent, but the main theorem leans on an unproved companion-paper input and there is a genuine p=2 gap in the secondary theorem. read the letter →

arxiv 1908.00607 v2 pith:G5P7K5ZP submitted 2019-08-01 math.AP math-phmath.MP

Shiwu Yang This is my paper · ORCID

classification math.APmath-phmath.MP MSC 35L7135B4035L05

keywords semilinearwaveequationpointwisedecayenergyscatteringweightedestimatesbackwardlightconeconformalcompactificationdefocusingnonlinearityvectorfieldmethod

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, for the defocusing energy-subcritical equation $\square\varphi=|\varphi|^{p-1}\varphi$ in $\mathbb{R}^{1+3}$, the pointwise decay of every global solution with finite weighted energy matches that of a free wave once $p>\frac{1+\sqrt{17}}{2}\approx 2.5616$. For the remaining range $22.3542$, a range that dips below the usual critical exponent $1+\sqrt{2}\approx 2.414$, and it proves uniform boundedness for $p>\frac{3}{2}$. The interest is that large nonlinearities do not destroy dispersion of waves: the nonlinear solution inherits the same decay away from the light cone as the linear solution, which is exactly what scattering requires.

What carries the argument

The carrying identity is the weighted null-flux estimate obtained by applying the multiplier $X^\gamma=v_+^\gamma(\partial_t+\partial_r)+u_+^\gamma(\partial_t-\partial_r)$, where $u_+=\sqrt{1+u^2}$, $v_+=\sqrt{1+v^2}$ in null coordinates $u=(t-r)/2$, $v=(t+r)/2$. With this multiplier and the energy identity on the past of a point $q$, the paper obtains a uniform bound on $(1+\tau)v_+^\gamma+u_+^\gamma$ times $|\varphi|^{p+1}$ integrated over the backward light cone, where $\tau$ is the cosine of the angle between the space directions; the potential term is controlled by a spacetime bound quoted from the companion paper. A second multiplier $r^\gamma(\partial_t+\partial_r)$ handles $p\le2$. Three integration lemmas turn these cone fluxes into pointwise control through the standard representation formula for the wave equation; the exterior region is treated directly, while the interior is mapped by conformal compactification to a truncated backward cone on which a bootstrap argument closes.

What would settle it

Construct a sequence of initial data with uniformly bounded $E_{0,\gamma_0}$ for which the weighted integral $\int\int v_+^{\gamma_0-1-\varepsilon}|\varphi|^{p+1}\,dx\,dt$ diverges as the data range expands; if such data exist, Proposition 3.1 fails and the pointwise decay theorems built on it cannot hold.

Watch

Extended reading notes

Core claim

At its core, the paper's claim is Theorem 1.1: for $p>\frac{1+\sqrt{17}}2$ and $\max\{\frac4{p-1}-1,1\}<\gamma_0<\min\{p-1,2\}$, any solution whose data lie in the weighted energy space $E_{1,\gamma_0}$ obeys $$|\varphi(t,x)|\le C\sqrt{E_{1,\gamma_0}}\,(1+t+|x|)^{-1}(1+||x|-t|)^{-(\gamma_0-1)/2}.$$ That is exactly the decay rate of the free wave with the same data, including the improved decay away from the light cone. For $2<p\le\frac{1+\sqrt{17}}2$ it obtains the slower but still explicit rate $(1+t+|x|)^{-\alpha_p\gamma_0}(1+||x|-t|)^{-\gamma_0/(p+1)}$ with $\alpha_p=\frac{3+(p-2)^2}{(p+1)(5-p)}$, and from these bounds it derives energy scattering for $p$ above a threshold $p_*$ with $2.3541<p_*<2.3542$. Theorem 1.2 adds that for $\frac32<p\le2$ the solution remains uniformly bounded in terms of the initial energy.

Load-bearing premise

Everything else rests on the unproved companion-paper bound that a certain weighted spacetime integral of $|\varphi|^{p+1}$—quoted here as Proposition 3.1—is controlled by the zeroth-order weighted energy; if that control fails, the decay and scattering results collapse.

Editorial extensions

If this is right

For $p>(1+\sqrt{17})/2$, the nonlinear solution satisfies the linear-wave decay $|\varphi(t,x)|\lesssim(1+t+|x|)^{-1}(1+||x|-t|)^{-(\gamma_0-1)/2}$, uniformly in space-time.
For $2<p\le(1+\sqrt{17})/2$, the pointwise decay is at least $t^{-1/3}$ in the worst case, with the explicit exponent $\frac{3+(p-2)^2}{(p+1)(5-p)}\gamma_0$ on $1+t+|x|$.
Energy scattering occurs for every $p>p_*$ with $p_*<2.3542$ and data in $E_{1,p-1}$; this goes below the usual critical power $1+\sqrt{2}\approx2.414$.
For $\frac32<p\le2$, global solutions are uniformly bounded by the initial weighted energy, even though time decay is not asserted.
The relevant weighted spacetime integral $\int v_+^{\gamma_0-1-\varepsilon}|\varphi|^{p+1}$ is finite, and this is the quantitative input that powers both the pointwise bounds and the scattering conclusion.

Reading between the lines

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

Inference: the two-region structure (exterior cone-flux bootstrap, interior conformal compactification) is modular; the same pairing is likely to transfer to curved backgrounds with well-separated null cones.
Inference: the threshold $p_*<2.3542$ is an artifact of the final mixed-norm interpolation; sharpening that step, or upgrading the quoted spacetime bound, could lower the scattering threshold further, possibly toward the usual critical value $2.414$.
Inference: the method's failure for $p\le2$ is caused by the sign condition on the weighted potential energy, not by an intrinsic obstruction; a different weight family may turn Theorem 1.2's boundedness into time decay across the whole range $p>\frac32$.

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

You should know two things about this paper. First, it is a real advance: linear-rate pointwise decay for p > (1+√17)/2, explicit algebraic decay for 2 < p ≤ (1+√17)/2, and energy scattering for p > 2.3542, below the Strauss exponent 1+√2. That beats Pecher’s scattering threshold of 2.7005 and the earlier pointwise decay range. The method is a careful, often elegant use of Dafermos–Rodnianski weighted multipliers plus a conformal compactification for the interior region. The cone-flux estimate in Proposition 3.2 is a genuinely nice piece of vector-field analysis. Second, the main result is not self-contained. Proposition 3.1, the uniform spacetime bound with the v_+^{γ0−1−ε} weight, is the load-bearing input for Theorem 1.1 and Corollary 1.1, and its proof is a one-line citation to the author’s companion paper [26], not yet published. The stress-test is right that if [26] is weaker, requires extra smallness, or has a different γ0 range, the decay picture collapses. This is not circular—[26] proves a spacetime bound, not the pointwise decay or scattering—but it is a heavy reliance on a to-appear self-citation. A referee cannot verify the core estimate without that companion paper. The secondary theorem also has a real p = 2 gap. In Section 7, the integral ∫₀^{t₀} (1+tilde r)^{−γ} tilde r^{1−p} dtilde r diverges logarithmically at 0 when p = 2, since γ = 2−p+ε becomes ε. The theorem as stated covers p = 2, so the proof does not close there. This does not affect Theorem 1.1 or Corollary 1.1, which require p > 2, but it is a genuine flaw in Theorem 1.2 and should be fixed by changing the endpoint or adding a separate estimate. My overall read: the central claims are probably correct, the combinatorial and conformal arguments are serious and mostly check out, and the improvements over known results are clear. But in its present form the paper deserves a major revision, not immediate acceptance. The author should make Proposition 3.1 available to the referee in a verifiable form—ideally by including the companion paper or a proof sketch with sufficient detail—and repair the p = 2 endpoint. If those are fixed, this becomes a solid paper. Send it to a serious referee, but with the requirement that the companion paper be part of the review package. I would not cite it yet in my own work until [26] is actually available.

Referee Report

3 major / 4 minor

Summary. The paper studies global solutions of the defocusing semilinear wave equation, □φ = |φ|^{p−1}φ, in R^{1+3} for 1 < p < 5. The main results are: (i) Theorem 1.1, which gives pointwise decay matching the linear wave rate, |φ(t,x)| ≲ (1+t+|x|)^{-1}(1+||x|−t|)^{-(γ0−1)/2}, for p > (1+√17)/2 under a weighted E_{1,γ0} norm; (ii) a slower decay rate for 2 < p ≤ (1+√17)/2; (iii) Corollary 1.1, energy-space scattering for p > p* with p* < 2.3542; and (iv) Theorem 1.2, uniform boundedness for 3/2 < p ≤ 2. The method uses Dafermos–Rodnianski vector fields, weighted flux estimates through backward light cones, a conformal compactification argument for the interior region, and bootstrap or Gronwall arguments in the exterior region.

Significance. If the proof holds, this is a genuine improvement over Pecher's decay estimates and extends energy scattering to powers below the Strauss exponent, which would be an important advance in the asymptotic theory of defocusing semilinear wave equations. The paper contains detailed multiplier computations, explicit decay rates, and a clear decomposition into exterior and interior regions. However, the central decay theorem depends on a uniform spacetime bound (Proposition 3.1) that is quoted from the author's companion paper [26] without proof in the present text; since that bound controls the sign-indefinite potential term and its exponent γ0 directly enters the final rates, the main result is conditional on an unverified external statement. In addition, the proof of Theorem 1.2 fails at the endpoint p = 2 because of a divergent integral. These are load-bearing gaps rather than presentation issues.

major comments (3)

[Section 3, Proposition 3.1, Eq. (6)] Proposition 3.1 is the sole input for controlling the sign-indefinite potential term in Proposition 3.2, and through it all later estimates in Sections 4–6 and both parts of Theorem 1.1. Its proof is only 'See the main theorem in [26]', and the precise range 1 < γ0 < min{2, p−1} together with the constant CE0,γ0 is inherited by all subsequent decay rates. If the companion paper's theorem has weaker hypotheses, requires small data, or has a different exponent range, the main theorem collapses. The present manuscript should either include a proof of Proposition 3.1 or state and prove a self-contained version with all hypotheses and the exact dependence on E0,γ0.
[Section 7, proof of Theorem 1.2, endpoint p = 2] The estimate for the nonlinear term concludes with ∫_0^{t0} (1+tilde r)^{-γ} tilde r^{1-p} dtilde r ≲ 1 + (1+t0)^{2-p-γ}. For p = 2 this integral diverges logarithmically at tilde r = 0 for every γ > 0, so the displayed bound is false at the endpoint. Since Theorem 1.2 includes p = 2, the stated proof does not close; the p = 2 case needs a separate truncation or limiting argument, or the theorem should exclude p = 2.
[Section 6, proof of Corollary 1.1] The numerical claim p* < 2.3542 rests on the function f(p), but the displayed formula in the text is garbled: 'f (p) = p − 2 + (p − 1)2 3 + (p − 2)2(5 − p)(p + 1) − 1' is not a well-formed expression. The intended definition appears to be f(p) = p − 3 + (p − 1)^2(3 + (p − 2)^2)/((5 − p)(p + 1)), and the monotonicity/root argument should be written out clearly, since this is the quantitative basis for the claimed scattering threshold.

minor comments (4)

[Throughout] The text contains many typographical and OCR artifacts, such as the title spacing 'W A VE EQUA TIONS' and the malformed formula for f(p); these should be corrected in the final version.
[Section 2] The null coordinate u is defined by 'u = t−r /2', which should be u = (t − r)/2; the absence of parentheses makes the normalization ambiguous, though the later use of u+ = sqrt(1 + u^2) indicates the intended definition.
[Section 3, proof of Proposition 3.2] Near the end of the proof the bound is written as '≤ CE0,γ+ǫ' and then one 'lets 0 < ǫ < γ0 − γ'; this is correct only because E0,γ+ǫ ≤ E0,γ0, and that monotonicity should be stated explicitly.
[Section 4, Eq. (15)] The function M(t0) is defined with a power ((p+1)/(2−ε_p)) of sup |u^{(γ0−1)/2} r φ|, and the notation is reused later for M1(t0); a brief note distinguishing these quantities would improve readability.

Circularity Check

0 steps flagged · score 2.0 of 10

No by-construction circularity; the main theorem's key spacetime bound is imported by self-citation from the author's companion paper [26], which is a correctness burden rather than a circular reduction.

full rationale

The derivation leading to Theorem 1.1 is not circular: Proposition 3.1, eq. (6), is a weighted spacetime L^{p+1} bound imported from the author's companion paper [26], and it is strictly a different, weaker statement than the pointwise decay or energy-scattering conclusions of Theorem 1.1 and Corollary 1.1. The subsequent argument—multiplier identities in Proposition 3.2, integration lemmas in Sections 4 and 5, the conformal compactification in Section 6, and the final bootstrap inequalities—does genuine work to extract pointwise decay from that bound. The decay rates are not fitted to the target, and the linear-wave comparison in Remark 1.3 is a benchmark, not an input. The main non-circular concern is self-containment: the proof of Proposition 3.1 is only 'See the main theorem in [26]', so the quantitative decay picture collapses if that companion theorem fails; this raises the burden but is a correctness/reproducibility risk, not a circular definition. A separate correctness gap, also not circular, is the p=2 endpoint of Theorem 1.2: in Section 7 the integral ∫_0^{t0}(1+rtilde r)^{-γ} rtilde r^{1-p} drtilde r is divergent logarithmically at rtilde r=0 when p=2, so the stated proof does not close at that endpoint.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no empirically fitted constants or invented entities. Its main input is a spacetime bound taken from the same author's companion paper [26], which acts as a domain-level assumption rather than a standard lemma with proof in this text. Auxiliary parameters epsilon, gamma, and beta are chosen small and are part of the proof, not free parameters.

assumptions (4)

standard math Global well-posedness and energy identity for the energy-subcritical defocusing NLW.
Cited to [3,4,11,13,15,17,20,24]; used to justify the energy identities and the flux integrals in Sections 3 and 7.
domain assumption Spacetime bound of Proposition 3.1: integral integral v_+^{gamma0-1-epsilon}|phi|^{p+1} <= C E0,gamma0.
Quoted without proof from the author's companion paper [26]; it is the starting point for the weighted cone flux bound (7) and for all later decay estimates.
standard math Representation formula (11) for the wave equation in R^{1+3}.
Used throughout Sections 4 and 7 to convert flux bounds into pointwise bounds; classical Kirchhoff formula.
standard math Conformal compactification maps the region D inside the hyperboloid H to a truncated backward light cone, preserving null hypersurfaces.
The explicit map and measure transforms are computed in Section 6; the conformal transformation of the nonlinearity uses the weight Lambda^{3-p}.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Pointwise decay for semilinear wave equations in $\mathbb{R}^{!+3}$." pith.science (2026). https://pith.science/paper/G5P7K5ZP

@misc{pith2026190800607,
  author       = {Pith},
  title        = {Pith review of: Pointwise decay for semilinear wave equations in $\mathbbR^!+3$},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G5P7K5ZP}},
  note         = {Machine review of arXiv:1908.00607}
}

abstract

In this paper, we use Dafermos-Rodnianski's new vector field method to study the asymptotic pointwise decay properties for solutions of energy subcritical defocusing semilinear wave equations in $\mathbb{R}^{1+3}$. We prove that the solution decays as quickly as linear waves for $p>\frac{1+\sqrt{17}}{2}$, covering part of the sub-conformal case, while for the range $2<p\leq \frac{1+\sqrt{17}}{2}$, the solution still decays with rate at least $t^{-\frac{1}{3}}$. As a consequence, the solution scatters in energy space when $p>2.3542$. We also show that the solution is uniformly bounded when $p>\frac{3}{2}$.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages

[26]

S. Yang. Global behaviors for defocusing semilinear wa ve equations. Ann. Sci. ´Ec. Norm. Sup´ er. (4). to appear. Beijing International Center for Mathematical Research, Peking University, Beijing, China Email: shiwuyang@bicmr.pku.edu.cn

work page
[1]

Large data pointwise decay for defocusing semilinear wave equations

R. Bieli and N. Szpak. Large data pointwise decay for defo cusing semilinear wave equations. 2010. arXiv:1002.3623

work page Pith review arXiv 2010
[2]

P. Brenner. On Lp − Lp′ estimates for the wave-equation. Math. Z. , 145(3):251–254, 1975

work page 1975
[3]

P. Brenner. On the existence of global smooth solutions o f certain semilinear hyperbolic equations. Math. Z. , 167(2):99– 135, 1979

work page 1979
[4]

Brenner and W

P. Brenner and W. von W ahl. Global classical solutions of nonlinear wave equations. Math. Z. , 176(1):87–121, 1981

work page 1981
[5]

Choquet-Bruhat and D

Y. Choquet-Bruhat and D. Christodoulou. Existence of gl obal solutions of the Yang-Mills, Higgs and spinor ﬁeld equations in 3 + 1 dimensions. Ann. Sci. ´Ecole Norm. Sup. (4) , 14(4):481–506 (1982), 1981

work page 1982
[6]

Christodoulou

D. Christodoulou. Global solutions of nonlinear hyperb olic equations for small initial data. Comm. Pure Appl. Math. , 39(2):267–282, 1986

work page 1986
[7]

Dafermos and J

M. Dafermos and J. Luk. The interior of dynamical vacuum b lack holes I:The C0-stability of the Kerr Cauchy horizon

work page

Show all 26 references

[8]

Dafermos and I

M. Dafermos and I. Rodnianski. A new physical-space appr oach to decay for the wave equation with applications to black hole spacetimes. In XVIth International Congress on Mathematical Physics , pages 421–432. W orld Sci. Publ., Hackensack, NJ, 2010

2010
[9]

Eardley and V

D. Eardley and V. Moncrief. The global existence of Yang- Mills-Higgs ﬁelds in 4-dimensional Minkowski space. I. Loc al existence and smoothness properties. Comm. Math. Phys. , 83(2):171–191, 1982

1982
[10]

Eardley and V

D. Eardley and V. Moncrief. The global existence of Yang -Mills-Higgs ﬁelds in 4-dimensional Minkowski space. II. Completion of proof. Comm. Math. Phys. , 83(2):193–212, 1982

1982
[11]

Ginibre and G

J. Ginibre and G. Velo. The global Cauchy problem for the nonlinear Klein-Gordon equation. Math. Z. , 189(4):487–505, 1985

1985
[12]

Ginibre and G

J. Ginibre and G. Velo. Conformal invariance and time de cay for nonlinear wave equations. I, II. Ann. Inst. H. Poincar´ e Phys. Th´ eor., 47(3):221–261, 263–276, 1987

1987
[13]

Ginibre and G

J. Ginibre and G. Velo. The global Cauchy problem for the nonlinear Klein-Gordon equation. II. Ann. Inst. H. Poincar´ e Anal. Non Lin´ eaire, 6(1):15–35, 1989

1989
[14]

F. John. Blow-up of solutions of nonlinear wave equatio ns in three space dimensions. Manuscripta Math. , 28(1-3):235– 268, 1979

1979
[15]

J¨ orgens

K. J¨ orgens. Das Anfangswertproblem im Grossen f¨ ur eine Klasse nichtlinearer Wellengleichungen. Math. Z. , 77:295–308, 1961

1961
[16]

Lindblad and J

H. Lindblad and J. Sterbenz. Global stability for charg ed-scalar ﬁelds on Minkowski space. IMRP Int. Math. Res. Pap., pages Art. ID 52976, 109, 2006

2006
[17]

H. Pecher. Lp-Absch¨ atzungen und klassische L¨ osungen f¨ ur nichtlineare Wellengleichungen. I. Math. Z. , 150(2):159–183, 1976

1976
[18]

H. Pecher. Decay of solutions of nonlinear wave equatio ns in three space dimensions. J. Funct. Anal. , 46(2):221–229, 1982

1982
[19]

H. Pecher. Scattering for semilinear wave equations wi th small data in three space dimensions. Math. Z. , 198(2):277– 289, 1988

1988
[20]

I. Segal. Non-linear semi-groups. Ann. of Math. (2) , 78:339–364, 1963

1963
[21]

W. Strauss. Decay and asymptotics for cmu = F (u). J. Functional Analysis , 2:409–457, 1968. POINTWISE DECAY FOR SEMILINEAR W A VE EQUATIONS IN R1+3 37

1968
[22]

Ettore Majorana

W. Strauss. Nonlinear invariant wave equations. In Invariant wave equations (Proc. “Ettore Majorana” Interna t. School of Math. Phys., Erice, 1977) , volume 73 of Lecture Notes in Phys. , pages 197–249. Springer, Berlin-New York, 1978

1977
[23]

von W ahl

W. von W ahl. Some decay-estimates for nonlinear wave eq uations. J. Functional Analysis , 9:490–495, 1972

1972
[24]

von W ahl

W. von W ahl. ¨Uber nichtlineare Wellengleichungen mit zeitabh¨ angigem elliptischen Hauptteil. Math. Z. , 142:105–120, 1975

1975
[25]

W ei and S

D. W ei and S. Yang. On the defocusing semilinear wave in t hree space dimension with small power. preprint

Pith tools

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

[1] [26]

S. Yang. Global behaviors for defocusing semilinear wa ve equations. Ann. Sci. ´Ec. Norm. Sup´ er. (4). to appear. Beijing International Center for Mathematical Research, Peking University, Beijing, China Email: shiwuyang@bicmr.pku.edu.cn

work page

[2] [1]

Large data pointwise decay for defocusing semilinear wave equations

R. Bieli and N. Szpak. Large data pointwise decay for defo cusing semilinear wave equations. 2010. arXiv:1002.3623

work page Pith review arXiv 2010

[3] [2]

P. Brenner. On Lp − Lp′ estimates for the wave-equation. Math. Z. , 145(3):251–254, 1975

work page 1975

[4] [3]

P. Brenner. On the existence of global smooth solutions o f certain semilinear hyperbolic equations. Math. Z. , 167(2):99– 135, 1979

work page 1979

[5] [4]

Brenner and W

P. Brenner and W. von W ahl. Global classical solutions of nonlinear wave equations. Math. Z. , 176(1):87–121, 1981

work page 1981

[6] [5]

Choquet-Bruhat and D

Y. Choquet-Bruhat and D. Christodoulou. Existence of gl obal solutions of the Yang-Mills, Higgs and spinor ﬁeld equations in 3 + 1 dimensions. Ann. Sci. ´Ecole Norm. Sup. (4) , 14(4):481–506 (1982), 1981

work page 1982

[7] [6]

Christodoulou

D. Christodoulou. Global solutions of nonlinear hyperb olic equations for small initial data. Comm. Pure Appl. Math. , 39(2):267–282, 1986

work page 1986

[8] [7]

Dafermos and J

M. Dafermos and J. Luk. The interior of dynamical vacuum b lack holes I:The C0-stability of the Kerr Cauchy horizon

work page