Pith. sign in

REVIEW 1 major objections 6 minor 23 references

Homogenization of linear parabolic equations with three spatial and three temporal scales for certain matchings between the microscopic scales

T0 review · 1 major / 6 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 a parabolic equation with two fast spatial and two fast temporal scales, the paper proves a homogenized limit that is elliptic, with resonance occurring only under a p-shifted matching of scales.

desk verdict A legitimate extension of the multiscale toolbox to three spatial and three temporal scales, with solid compactness theorems and a plausible 13-case homogenization classification, but the proof of Theorem 10 skips the cell-problem well-posedness and coercivity checks that the uniqueness statement needs. read the letter →

arxiv 1908.05892 v1 pith:2LLBS3HE submitted 2019-08-16 math.AP

Tatiana Danielsson , Pernilla Johnsen This is my paper · ORCID
classification math.AP MSC 35B2735K20
keywords homogenizationmultiscaleconvergenceveryweakparabolicequationselliptichomogenizedproblemresonanceperiodiccoefficientscellproblems
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 establishes a homogenization limit for a linear parabolic equation with two rapidly oscillating spatial scales and two rapidly oscillating temporal scales (three spatial and three temporal scales in total, counting the macroscopic ones): $\varepsilon^{p}\partial_{t}u_{\varepsilon}-\nabla\cdot(a(x/\varepsilon,x/\varepsilon^{2},t/\varepsilon^{q},t/\varepsilon^{r})\nabla u_{\varepsilon})=f$, with $0

What carries the argument

The machinery has three parts. First, evolution multiscale convergence and very weak multiscale convergence (Definitions 2 and 7) replace classical two-scale convergence for sequences with two spatial and two temporal microscopic scales. Second, Theorem 6 characterizes the $(2,3)$- and $(3,3)$-scale limits of $\nabla u_{\varepsilon}$ under the integral conditions (2)–(3), conditions that stand in for boundedness of $\partial_{t}u_{\varepsilon}$ in $L^{2}(0,T;H^{-1})$, and Theorem 9 identifies the very weak limits of $\varepsilon^{-1}u_{\varepsilon}$ and $\varepsilon^{-2}u_{\varepsilon}$ with the same correctors $u_1$ and $u_2$. Third, in the homogenization proof, two families of test functions, (59) and (61), with adjustable powers $k$ of $\varepsilon$, are inserted into the weak formulation; choosing $k=r-p-2$, $q-p-2$, $r-p-1$, or $q-p-1$ isolates each local time variable and produces the thirteen systems of local problems that define the homogenized coefficient $b$.

What would settle it

Set $N=1$, $p=1$, $q=3$ (so $q=2+p$), $r=4$, and choose the coefficient $a(y_2,s_2)=2+\sin(2\pi s_2)$. Solve the two local problems of Theorem 10, case 4, to get the predicted homogenized coefficient $b$, then compute the actual effective coefficient from a two-scale asymptotic expansion of (18); agreement would confirm the $p$-shifted resonance, while a mismatch would indicate the matching condition is wrong.

Watch

Extended reading notes

Core claim

The central claim is Theorem 10: for every fixed $\varepsilon$, the parabolic problem (18) has a unique solution, and as $\varepsilon\to 0$ the solutions converge weakly in $L^{2}(0,T;H_{0}^{1}(\Omega))$ to the unique solution of the elliptic homogenized problem $-\nabla\cdot(b\nabla u)=f$ with $u=0$ on the boundary. The gradient has the three-scale decomposition $\nabla u_{\varepsilon}\rightharpoonup \nabla u+\nabla_{y_1}u_1+\nabla_{y_2}u_2$, where the correctors $u_1$ and $u_2$ solve local problems that are elliptic in most of the thirteen parameter regimes and parabolic only in the resonant ones. The two phenomena emphasized by the paper are therefore: the homogenized problem is elliptic, and the resonance condition is shifted by $p$ — a temporal scale $\varepsilon^{q}$ or $\varepsilon^{r}$ acts like the square of a spatial scale when $q-p=2$ or $r-p=2$, with the analogue $q-p=4$ or $r-p=4$ for the second spatial scale, rather than when $q=2$ or $r=2$ as in the $p=0$ case. Theorems 6 and 9 provide the underlying compactness: under the integral conditions (2)–(3), the $(3,3)$-scale limit of the gradient is characterized by $u,u_1,u_2$, and the unbounded sequences $\varepsilon^{-1}u_{\varepsilon}$ and $\varepsilon^{-2}u_{\varepsilon}$ converge in the very weak multiscale sense to the same correctors.

Load-bearing premise

The whole result rests on the strict ordering $0<p<q<r$ together with the two integral conditions (2)–(3), which stand in for the usual bound on the time derivative; if those fail, the limit could keep oscillating in the fast time variables and the elliptic homogenized equation would not follow.

Editorial extensions

If this is right

  • If the paper is right, the macroscopic behavior of this multiscale parabolic medium is instantaneous: the effective equation is elliptic, so the initial data do not appear in the leading-order limit problem.
  • The effective diffusion matrix $b$ is computed from cell problems that are elliptic except when $r=2+p$, $q=2+p$, $r=4+p$, or $q=4+p$, in which case one of the cell problems becomes parabolic.
  • The convergence statements give rigorous meaning to the correctors: $\varepsilon^{-1}u_{\varepsilon}$ and $\varepsilon^{-2}u_{\varepsilon}$ converge very weakly to the same $u_1,u_2$ that appear in the gradient decomposition.
  • The compactness results do not require a uniform bound on $\partial_t u_{\varepsilon}$ in $L^{2}(0,T;H^{-1})$; the integral conditions (2)–(3), verified directly from the equation, are enough.

Reading between the lines

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

  • The $p$-shift suggests a general rule: whenever the time derivative carries a vanishing factor $\varepsilon^{p}$, resonance between a spatial scale $\varepsilon^{k}$ and a temporal scale $\varepsilon^{s}$ should occur at $s-p=2k$; this could be tested with a more general vanishing factor $\varphi(\varepsilon)$ instead of $\varepsilon^{p}$.
  • Because the homogenized equation is elliptic, one expects an initial layer and boundary layers to carry the lost time information; a natural extension would construct correctors that capture these layers in stronger norms, which the paper does not do.
  • The very weak multiscale convergence of $\varepsilon^{-1}u_{\varepsilon}$ and $\varepsilon^{-2}u_{\varepsilon}$ shows that the correctors are uniquely determined even for unbounded sequences, suggesting the same condition-based compactness could handle stronger singular scalings if the test-function powers are adjusted accordingly.
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.

Referee Report

1 major / 6 minor

Summary. The paper establishes compactness results for evolution multiscale and very weak multiscale convergence for sequences bounded in L2(0,T;H1_0(Ω)) that satisfy the integral conditions (2) and (3), which replace the usual boundedness of the time derivative in L2(0,T;H^{-1}(Ω)). These results are then applied to homogenize the parabolic problem ε^p ∂t uε - ∇·(a(x/ε,x/ε^2,t/ε^q,t/ε^r)∇uε)=f with 0<p<q<r. The main homogenization theorem, Theorem 10, states that uε converges weakly in L2(0,T;H1_0(Ω)) to the unique solution u of the elliptic problem -∇·(b∇u)=f, with the homogenized coefficient b characterized by 13 families of local problems depending on the relative sizes of p,q,r. The paper highlights two phenomena: the homogenized problem is elliptic even though the original problem is parabolic, and parabolic resonance occurs when a temporal scale multiplied by ε^{-p} matches the square of a spatial scale.

Significance. If the gaps identified below are filled, the paper is a competent extension of the multiscale homogenization framework of Allaire-Briane and Flodén et al. The compactness theorems are of independent interest because they avoid the usual bound on the time derivative. The derivation is essentially self-contained and the nonstandard hypotheses (2)-(3) are verified directly from the equation in Section 3. The 13-case homogenization theorem gives concrete, falsifiable predictions for different scale matchings. The main weakness is that the proof of Theorem 10 does not supply the standard well-posedness and coercivity arguments on which the uniqueness and whole-sequence convergence claims rest; these arguments are routine and local, so the central claim is defensible.

major comments (1)
  1. [Section 3, Theorem 10 (after Eq. (62))] The theorem asserts that u is the unique solution of the elliptic homogenized problem (26) and that b is characterized by the local problems (27)-(57). The proof passes to the limit and derives weak forms of these local problems, but it never proves that the cell problems are well-posed or that b is coercive. For the parabolic cell problems, e.g. (31), (36), (40), (42)-(43), (44), (50) and (54), existence of a periodic-in-time solution requires a Fredholm compatibility condition; this condition is indeed automatic for the displayed equations because their right-hand sides are in divergence form, but that verification is not included. Coercivity of b, needed for the uniqueness assertion in (26), follows from standard energy identities obtained by testing the local equations with u1 and u2 and integrating by parts, but these identities are not stated anywhere. Because b must be shown to be a well-defined coercive tensor for the limit u to be unique and for the whole sequence (rather than a subsequence) to converge, this is a load-bearing gap. Please add a lemma (or a paragraph in the proof of Theorem 10) proving well-posedness of each type of local problem and the estimate bξ·ξ ≥ C0|ξ|^2 for all ξ ∈ R^N.
minor comments (6)
  1. [Section 3, Eq. (58)] In the displayed weak form (58), the first occurrence of 'u(x,t)' inside the integral should be '∇u(x,t)'; as printed, the expression is not the weak form of (26) and is dimensionally inconsistent.
  2. [Section 2, proof of Theorem 6] After the divergence-free test functions are introduced, the text invokes the H^{-1}(Ω)-boundedness of ε^{-2}∇y1·v and then passes to the limit. It should say explicitly that ε^{-1}∇y1·v = ε(ε^{-2}∇y1·v) → 0 in H^{-1}(Ω); the current wording leaves this step implicit.
  3. [Section 2, Theorem 6 and Theorem 9] The hypotheses of these theorems do not restate the joint well-separatedness of the scale lists {ε,ε^2} and {ε^q,ε^r}. This property is automatic for 0<q<r (after removing duplicates the exponents are strictly increasing), but stating it explicitly would make the theorems self-contained.
  4. [Theorem 10, cases 2, 4, 6, 7, 10, 12] The notation such as 'u1 ∈ L2(Ω_T × S1; W1,2)' is nonstandard and potentially confusing; a sentence in Notation 1 clarifying that this means u1(·,s2) ∈ W1,2 for almost every (x,t,s1) would remove ambiguity.
  5. [Introduction] The relationship to the authors' earlier works [13] and [6] could be made more explicit: [13] treats one spatial and one temporal scale, and the novelty of the two-scale spatial/temporal setting and of the shifted resonance condition should be stated more concretely.
  6. [References] Reference [5] is a preprint and [6] is 'to appear'; if published or updated versions exist, they should be cited with full bibliographic data.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular reduction found; the result is derived from the equation and standard external compactness theorems.

full rationale

The derivation chain is self-contained and exhibits no circular reduction. The compactness Theorems 6 and 9 are proved from the hypotheses (2)-(3) using standard multiscale tools from Allaire-Briane [2] and Flodén et al. [11]; in Section 3 the hypotheses (20)-(21) are verified from the weak form (22) of the original equation, not assumed as the target conclusion. The homogenized problem (26) and the thirteen characterizations of b are obtained by passing to the limit in the weak form with oscillating test functions (59) and (61) and applying Theorems 6 and 9; the local problems are the resulting limit equations, not inputs. Citations to the authors' earlier works [13] and [6] are contextual and are not load-bearing: the shifted-resonance and ellipticity phenomena are re-derived from the present equation. The skeptic's concern that well-posedness or coercivity of the local problems and of (26) is not proved is a correctness gap, not a circularity, because the theorem's statements are not assumed in the proof but asserted without full verification.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no free parameters and no new entities. It relies on the established multiscale convergence framework, with the main new structural input being the integral conditions (2)-(3) and their verification for the solution sequence.

assumptions (3)
  • standard math Compactness theorem for evolution multiscale convergence (Theorem 4) from [11], requiring jointly separated scales.
    Used at the start of the proof of Theorem 6 to pass to subsequential multiscale limits, equations (8)-(9).
  • standard math Allaire-Briane Lemma 3.7, the orthogonal decomposition of L^2 in reiterated homogenization, cited from [2].
    Used to characterize tau_0 - grad u as grad_y1 u1 + grad_y2 u2 in the proof of Theorem 6.
  • domain assumption Well-posedness of the local cell problems (elliptic and parabolic) in the stated spaces, including W_{i,j}, is assumed but not proved.
    Theorem 10 states the local problems and their solution spaces without existence proofs; existence and uniqueness are standard for coercive linear cell problems but are not demonstrated in the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Homogenization of linear parabolic equations with three spatial and three temporal scales for certain matchings between the microscopic scales." pith.science (2026). https://pith.science/paper/2LLBS3HE

@misc{pith2026190805892,
  author       = {Pith},
  title        = {Pith review of: Homogenization of linear parabolic equations with three spatial and three temporal scales for certain matchings between the microscopic scales},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2LLBS3HE}},
  note         = {Machine review of arXiv:1908.05892}
}
abstract

In this paper we establish compactness results of multiscale and very weak multiscale type for sequences bounded in $L^{2}(0,T;H_{0}^{1}(\Omega ))$, fulfilling a certain condition. We apply the results in the homogenization of the parabolic partial differential equation $\varepsilon ^{p}\partial_{t}u_{\varepsilon }\left( x,t\right) -\nabla \cdot \left( a\left( x/\varepsilon ,x/\varepsilon ^{2},t/\varepsilon^{q},t/\varepsilon ^{r}\right) \nabla u_{\varepsilon }\left( x,t\right)\right) = f\left( x,t\right) $, where $0<p<q<r$. The homogenization result reveals two special phenomena, namely that the homogenized problem is elliptic and that the matching for when the local problem is parabolic is shifted by $p$, compared to the standard matching that gives rise to local parabolic problems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 22 canonical work pages

  1. [1]

    Allaire : Homogenization and two-scale convergence

    G. Allaire : Homogenization and two-scale convergence. SIAM J. Math. Anal. 23 (1992), no. 6, 1482–1518. DOI 10.1137/0523084, Zbl 0770.3500 5, MR1185639

    work page arXiv 1992

  2. [2]

    Allaire, M

    G. Allaire, M. Briane : Multiscale convergence and reiterated homogeniza- tion. Proc. Roy. Soc. Edinburgh Sect. A. 126 (1996), no. 2, 297–342. DOI 10.1017/S0308210500022757, Zbl 0866.35017, MR1386865

    work page arXiv 1996

  3. [3]

    Allaire, A

    G. Allaire, A. Piatnitski : Homogenization of nonlinear reaction-diffusion equation with a large reaction term. Ann. Univ. Ferrara Sez. VII Sc i. Mat. 56 (2010), no. 1, 141–161. DOI 10.1007/s11565-010-0095-z, Zbl 1205.35019, MR2646529

    work page arXiv 2010

  4. [4]

    Bensoussan, J.-L

    A. Bensoussan, J.-L. Lions, G. Papanicolaou : Asymptotic analysis for periodic structures. Studies in Mathematics and its Applications 5, North-Holland Publishing Co., Amsterdam-New York, 1978. DOI 10.1090/chel/374, Zbl 0404.35001, MR0503330

    work page arXiv 1978

  5. [5]

    Danielsson, P

    T. Danielsson, P. Johnsen : Homogenization of the heat equation with a vanishing volumetric heat capacity. Preprint (2018). arXiv:1809.11 019

    work page 2018

  6. [6]

    Danielsson, P

    T. Danielsson, P. Johnsen : Homogenization of the heat equation with a vanishing volumetric heat capacity. Progress in Industrial Mathem atics at ECMI 2018, to appear. 26

    work page 2018

  7. [7]

    Douanla, J

    H. Douanla, J. L. Woukeng : Homogenization of reaction-diffusion equa- tions in fractured porous media. Electron J. Differential Equations . 2015 (2015), no. 253, 23 pp. Zbl 1336.35046, MR3414107

    work page arXiv 2015

  8. [8]

    Flod´ en, A

    L. Flod´ en, A. Holmbom, M. Olsson, J. Persson : Very weak multi- scale convergence. Appl. Math. Lett. 23 (2010), no. 10, 1170–1173. DOI 10.1016/j.aml.2010.05.005, Zbl 1198.35023, MR2665589

    work page arXiv 2010

Show all 23 references
  1. [9]

    Flod´ en, A

    L. Flod´ en, A. Holmbom, M. Olsson Lindberg: A strange term in the homog- enization of parabolic equations with two spatial and two temporal s cales. J. Funct. Spaces Appl. 2012 (2012), 9 pp. DOI 10.1155/2012/643458, Zbl 1242.35030, MR2875184

    2012

  2. [10]

    Flod´ en, A

    L. Flod´ en, A. Holmbom, M. Olsson Lindberg, J. Persson : A note on parabolic homogenization with a mismatch between the spatial scales . Abstr. Appl. Anal. 2013 (2013), 6 pp. DOI 10.1155/2013/329704, Zbl 1293.35027, MR3111807

    2013

  3. [11]

    Flod´ en, A

    L. Flod´ en, A. Holmbom, M. Olsson Lindberg, J. Persson : Homogenization of parabolic equations with an arbitrary number of scales in both spa ce and time. J. Appl. Math. 2014 (2014), 16 pp. DOI 10.1155/2014/101685, MR3176810

    2014 doi

  4. [12]

    Holmbom : Homogenization of parabolic equations: an alternative ap- proach and some corrector-type results

    A. Holmbom : Homogenization of parabolic equations: an alternative ap- proach and some corrector-type results. Appl. Math. 42 (1997), no. 5, 321–

    1997

  5. [13]

    Johnsen, T

    P. Johnsen, T. Lobkova : Homogenization of a linear parabolic problem with a certain type of matching between the microscopic scales. App l. Math. 63 (2018), no. 5, 503–521. DOI 10.21136/AM.2018.0350-17, Zbl 069 86923, MR3870146

    2018

  6. [14]

    Lobkova : Homogenization of linear parabolic equations with a certain resonant matching between rapid spatial and temporal oscillations in peri- odically perforated domains

    T. Lobkova : Homogenization of linear parabolic equations with a certain resonant matching between rapid spatial and temporal oscillations in peri- odically perforated domains. Acta Math. Appl. Sin. Engl. Ser. 35 (2019), no. 2, 340–358. DOI 10.1007/s10255-019-0810-1, MR3950176

    2019 doi

  7. [15]

    Lukkassen, G

    D. Lukkassen, G. Nguetseng, P. Wall : Two-scale convergence. Int. J. Pure Appl. Math. 2 (2002), no. 1, 35–86. Zbl 1061.35015, MR1912819

    2002

  8. [16]

    Nguetseng : A general convergence result for a functional related to the theory of homogenization

    G. Nguetseng : A general convergence result for a functional related to the theory of homogenization. SIAM J. Math. Anal. 20 (1989), no. 3, 608–623. DOI 10.1137/0520043, Zbl 0688.35007, MR0990867

    1989

  9. [17]

    Nguetseng : Asymptotic analysis for a stiff variational problem arising in mechanics

    G. Nguetseng : Asymptotic analysis for a stiff variational problem arising in mechanics. SIAM J. Math. Anal. 21 (1990), no. 6, 1394–1414. DOI 10.1137/0521078, Zbl 0723.73011, MR1075584 27

    1990

  10. [18]

    Nguetseng, J

    G. Nguetseng, J. L. Woukeng : Σ-convergence of nonlinear parabolic operators. Nonlinear Anal. 66 (2007), no. 4, 968–1004. DOI 10.1016/j.na.2005.12.035, Zbl 1116.35011, MR2288445

    2007

  11. [19]

    Persson : Selected Topics in Homogenization

    J. Persson : Selected Topics in Homogenization. Mid Sweden University Doctoral Thesis 127, Department of Engineering and Sustainable D evelop- ment, Mid Sweden University, 2012

    2012

  12. [20]

    Persson : Homogenization of monotone parabolic problems with sev- eral temporal scales

    J. Persson : Homogenization of monotone parabolic problems with sev- eral temporal scales. Appl. Math. 57 (2012), no. 3, 191–214. DOI 10.1007/s10492-012-0013-z, Zbl 1265.35018, MR2984600

    2012

  13. [21]

    Svanstedt, J

    N. Svanstedt, J. L. Woukeng : Periodic homogenization of strongly non- linear reaction-diffusion equations with large reaction terms. Appl. A nal. 92 (2013), no. 7, 1357–1378. DOI 10.1080/00036811.2012.678334, Zbl 1271.35006, MR3169106

    2013

  14. [22]

    Zeidler : Nonlinear functional analysis and its applications II/A: linear monotone operators

    E. Zeidler : Nonlinear functional analysis and its applications II/A: linear monotone operators. Springer-Verlag, New York, 1990. DOI 10.1 007/978- 1-4612-0985-0, Zbl 0684.47028, MR1033497 28

    1990

  15. [343]

    DOI 10.1023/A:1023049608047, Zbl 0898.35008, MR1467553

Pith tools

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