Pith. sign in

REVIEW 3 major objections 5 minor 26 references

H\"{o}lder continuous weak solutions of the 3D Boussinesq equation with thermal diffusion

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves that for every Hölder exponent below 1/3, the 3D Boussinesq equation with thermal diffusion has distributional weak solutions with a prescribed kinetic energy and a temperature energy identity.

desk verdict The target is right and the architecture is standard, but the temperature estimate that closes the induction undercounts derivatives in (7.22), so as written the proof does not close. read the letter →

arxiv 2506.02927 v1 pith:ICIOJ35F submitted 2025-06-03 math.AP

Zipeng Chen , Zhaoyang Yin This is my paper
classification math.AP MSC 35Q3076D03
keywords BoussinesqequationHöldercontinuousperiodicweaksolutionsConvexintegrationOnsagerexponentThermaldiffusionTransport-diffusionPrescribedkineticenergyMikadoflows
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proves that the three-dimensional Boussinesq equation with thermal diffusion admits Hölder continuous weak solutions at every spatial regularity exponent below 1/3, the Onsager-critical exponent for the incompressible Euler equations. For any smooth, strictly positive prescribed kinetic energy e(t) and any smooth initial temperature that depends only on the vertical coordinate and has zero mean, the constructed pair (v, θ) solves the system in the distribution sense, preserves the prescribed kinetic energy, and satisfies the temperature energy identity. If the construction is correct, the thermal coupling and diffusion do not lower the threshold reachable by convex integration, generalizing the Euler-case solution of Onsager's conjecture to a coupled active-scalar system. The significance is that earlier Boussinesq constructions with thermal effects achieved lower exponents, so this result brings the diffusive Boussinesq system into the same critical-regularity class as Euler.

What carries the argument

The proof is built on the convex integration framework for Onsager's conjecture: mollify, glue with the inverse-divergence operator R, then add a divergence-free perturbation w_{q+1} made of Mikado flows (high-frequency, zero-mean vector fields parameterized by a Reynolds tensor), so that v_{q+1} = v̄_q + w_{q+1} solves the Boussinesq–Reynolds equation with a small Reynolds stress. The new ingredient is the temperature: given v_{q+1}, the next temperature θ_{q+1} is defined by solving the transport-diffusion equation with initial data θ0(x3). The decisive estimate is Proposition 7.3, which bounds ‖θ_{q+1} − θ_q‖_{L2} by C ($δ_q^{{1/2}}$ λ_q)^α $l^{{1−α}}$; this is obtained by splitting the difference into three terms and applying the oscillatory transport-diffusion lemma (Lemma C.2). Together with a parameter inequality on the growth rate b, that bound makes the temperature-induced Reynolds stress I2 small enough to close the induction for any α < 1/3.

What would settle it

A concrete check is to compute ‖θ_{q+1} − θ_q‖_{L2} for the first few steps of the constructed sequence and verify it decays at the rate prescribed by (7.9); a counterexample would be a choice of smooth positive e(t) and smooth zero-mean θ0(x3) for which (7.9) or the parameter inequality (8.7) fails, so that the induction cannot close.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for any smooth strictly positive e(t) and any smooth θ0(x3) with zero mean, and for any α ∈ (0, 1/3), there exist v ∈ C^α([0,T] × $T^{3}$) and θ ∈ $C_t^{{1,α/2}}$ $C_x^{{2,α}}$ that solve the 3D Boussinesq system with thermal diffusion in the sense of distributions, satisfy the prescribed kinetic energy e(t) = ∫_{$T^{3}$} |v(t,x)|^2 dx, and satisfy the temperature energy identity. The proof also yields an Onsager-type dichotomy for the coupled system: for exponents above 1/3 the quantities E(t) and M(t) are conserved, while for exponents below 1/3 the kinetic energy E(t) can be prescribed to be non-constant while M(t) remains constant. If the theorem is true, the diffusive Boussinesq system belongs to the same Onsager-critical regularity class as the Euler equations.

Load-bearing premise

The load-bearing premise is the new L2 estimate (7.9) for the temperature difference, together with the structural assumption that the initial temperature depends only on x3 and has zero mean; if either fails, the induction closure and the applicability of the inverse-divergence operator R in the gluing step break down.

Editorial extensions

If this is right

  • For every α < 1/3, there exist Hölder continuous weak solutions of the 3D Boussinesq system with thermal diffusion that dissipate or gain kinetic energy according to a prescribed profile e(t).
  • The Onsager dichotomy holds for the coupled system: above 1/3, E(t) and M(t) are conserved; below 1/3, E(t) can be prescribed while M(t) stays constant.
  • The temperature inherits parabolic space-time regularity C_t^{1,α/2} C_x^{2,α} from the transport-diffusion equation, matching the Hölder exponent of the velocity.
  • Choosing the zero initial temperature recovers the Euler construction as a special case, so the theorem strictly extends the Onsager-critical result to the coupled system.

Reading between the lines

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

  • Editorial inference: the same scheme may extend to other active-scalar couplings, such as fractional dissipation in the temperature equation, as long as an oscillatory transport-diffusion estimate analogous to Lemma C.2 remains valid.
  • Editorial inference: the restriction that θ0 depends only on x3 with zero mean looks technical—it preserves the mean-free property needed for the inverse-divergence operator R; a mean-corrected gluing step could potentially admit arbitrary θ0.
  • Editorial inference: reaching the 1/3 threshold despite the coupling suggests that quadratic velocity–temperature transport does not change the Onsager critical exponent for the velocity in three dimensions, and one might test this by attempting the same construction with a velocity-dependent forcing.
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.

Referee Report

3 major / 5 minor

Summary. The paper constructs, for any Hölder exponent below 1/3, continuous weak solutions (v, θ) of the 3D Boussinesq system with thermal diffusion on the torus, with a prescribed positive kinetic energy e(t) and an exact temperature energy balance, for initial temperatures depending only on x3 with zero mean. The proof is a convex-integration induction in the style of Buckmaster–De Lellis–Székelyhidi–Vicol [5]: starting from a Boussinesq–Reynolds system, the authors mollify, glue exact solutions, add a Mikado perturbation w_{q+1}, solve the transport–diffusion equation for the new temperature θ_{q+1}, and then define the new Reynolds stress as I1 + I2 + I3. The central new ingredient is the L2 estimate (7.9) for θ_{q+1} − θ_q, which is used to bound the temperature-induced part I2 of the Reynolds stress. The paper is clearly written and carefully tracks the many parameters, but the proof of (7.9) rests on an oscillatory-phase derivative bound that appears incorrect as stated.

Significance. If the main theorem is correct, it is a substantial advance: it reaches the Onsager-critical range α < 1/3 for the 3D Boussinesq system with thermal diffusion, improves on previous results at regularity 1/5, and extends the 2D construction of [21] while also prescribing the kinetic energy and preserving the temperature energy identity. The constructive framework is standard and the paper is honest in attributing many technical ingredients to [5] and [21]; there is no apparent circularity, since the kinetic energy is prescribed a priori and the solution is built inductively. However, the proof as written does not establish the key temperature estimate: the bound (7.22) on the oscillatory phase is internally inconsistent with the flow-map estimates used elsewhere in the paper. The significance is therefore conditional on a correct replacement for Proposition 7.3 and on supplying the deferred proofs of several load-bearing estimates.

major comments (3)
  1. [Section 7, Eq. (7.22)] The bound (7.22) is not valid, and the proof of Proposition 7.3 does not close. For N = 1, the composition estimate (A.3) together with (6.10) gives ∥e^{iλ_{q+1}k·(Φ_i−x)}∥_1 ≲ λ_{q+1}∥∇Φ_i − Id∥_0 ≲ λ_{q+1} τ_q δ_q^{1/2} λ_q = λ_{q+1} l^{2α}, which is not O(|k|). More generally, the Faà di Bruno expansion contains the term λ_{q+1}∥D^N(Φ_i − x)∥_0, and (6.11) bounds ∥D^N Φ_i∥_0 only by l^{−N+1}; with l as in (4.1) this term is of order λ_q^{b + (N−1)(bβ − β + 1 + 3α/2)}, whereas the asserted λ_{q+1}^{N−1} is λ_q^{b(N−1)}. The discrepancy in the exponent is positive; for N = 1 it is b > 0. Consequently the estimate ∥∇^N d_{i,k}∥_{L2} ≲ δ_{q+1}^{1/2} λ_{q+1}^{N−1} used in (7.23) is not established. The second term of (7.23) is at best of order δ_{q+1}^{1/2}, not δ_{q+1}^{1/2} λ_{q+1}^{−1}, and the parameter inequality (8.7) cannot absorb that term. Since (7.9) is the only input that controls the temperature-induced stress I2 in (8.5), this is a load-bearing gap in the induction.
  2. [Section 8, Eq. (8.3)] The main Reynolds stress bound ∥I1∥_0 ≤ (1/3) δ_{q+2} λ_{q+1}^{−3α} is asserted without proof, with only the remark 'here we omit the proof' and a reference to [5, Proposition 6.1]. This is a load-bearing estimate: I1 is one of the three summands of the new Reynolds stress, and the induction requires exactly the powers stated in (8.3). The present construction differs from [5] through the temperature force θ_l e3 in the exact solutions (5.2), the definition of ρ_{q,i} involving e(t), and the modified Reynolds stress R_{q,i}. The estimate is therefore not a verbatim special case of [5]. Please provide a proof of (8.3) or a precise reduction showing that every term in (8.1) satisfies the same bounds as in [5].
  3. [Section 6, Proposition 6.8] The energy estimate δ_{q+2} λ_{q+1}^{−α} ≤ e(t) − ∫_{T3} |v_{q+1}|^2 ≤ δ_{q+2} is cited from [5, Proposition 6.2] without proof. This proposition is exactly what yields the prescribed kinetic energy (1.2) after passing to the limit, so it is central to Theorem 1.1. The proof in [5] relies on the particular form of ρ_q, the partition of unity in time, and the cancellation in the definition of the pressure; the authors should either reproduce the argument or state explicitly which hypotheses of [5, Proposition 6.2] are unchanged and why the temperature coupling and the new starting data do not affect it.
minor comments (5)
  1. [Throughout] The notation α is used both for the target Hölder exponent in Theorem 1.1 and for the auxiliary small parameter in the construction, e.g., in (2.4), (4.1), and Proposition 2.1. These are different objects and should be renamed to avoid confusion.
  2. [Abstract and Theorem 1.1] The abstract omits the hypotheses on θ0 (smooth, depending only on x3, zero mean) and the temperature energy identity (1.3); the statement of the main theorem should be reflected in the abstract.
  3. [Section 7, Proposition 7.1] In the proof of (7.5), the displayed estimate for θ_q uses δ_{q+1}^{1/2} λ_{q+1}^{k} in the intermediate sum, although θ_q is driven by v_q and should involve δ_q and λ_q. This appears to be an index error, and the final bound should read δ_q^{1/2} λ_q^{N−1}.
  4. [Appendix C, Lemma C.2] Lemma C.2 is stated for |k| = 1 but is applied to arbitrary k ∈ Z^3 \ {0}. The intended normalization λ|k| appears in (7.23), but the lemma should state explicitly that the frequency is λ|k| for general k.
  5. [General] There are minor typographical issues, including 'apprroximate' in the abstract and the inconsistent rendering of 'Hölder'; these should be corrected in a revision.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the prescribed kinetic energy is an input to the convex-integration ansatz, and all load-bearing estimates are either proved in the paper or cited to external prior work; no self-citation chain or fitted prediction reduces the main theorem to its own assumptions.

full rationale

The paper is an inductive convex-integration existence proof. The target energy e(t) is inserted at the base step and in the perturbation amplitude, not extracted from the output: in Section 3 the starting velocity v0 is chosen with amplitude containing 2e(t)-delta_1-delta_1 lambda_0^{-alpha}, and in Section 6.2 the perturbation is defined through rho_q(t) = (1/3)(e(t)-delta_{q+2}/2 - integral |vbar_q|^2 dx). The energy identity e(t)=integral |v|^2 dx then follows from the construction by Proposition 6.8, so this is a prescribed-energy existence result rather than a fitted quantity renamed as a prediction. The temperature increment bound (7.9) is derived by solving the transport-diffusion equation for theta_{q+1} with the previously constructed velocity v_{q+1}, and the proof decomposes theta_{q+1}-theta_q into f1,f2,f3 that solve forced linear parabolic equations with zero initial data; the estimate is not assumed as the definition of theta_{q+1}. The inverse-divergence operator R is used only after Remark 5.4 proves vi-vi+1 has zero mean using the zero-mean assumption on theta0; that is a stated hypothesis of Theorem 1.1, not a conclusion imported from the desired result. All borrowed results ([1], [5], [12], [13], [21], and classical Schauder estimates) are from sources external to the present authors, and no self-citation is load-bearing. The skeptic's highlighted estimate (7.22) concerns the size of derivatives of the oscillatory phase; even if that estimate were incorrect, it would be a technical correctness issue about lambda-decay in Proposition 7.3, not a circular reduction, because the estimate is asserted from derivative-counting computations rather than borrowed from the conclusion being proved. The derivation chain is self-contained in the circularity sense, and the paper is not using its own output as an input.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities or fitted constants. The proof depends on standard convex integration machinery and classical estimates. The auxiliary parameters a, b, alpha, s are chosen inside the proof to satisfy inequalities; their existence is part of the construction. The principal specialized input is the class of initial temperatures theta0(x3) with zero mean.

free parameters (4)
  • a = sufficiently large (a tends to infinity)
    Geometric parameter controlling lambda_q = 2 pi ceil(a^{b^q}); all estimates require a large enough to absorb constants and satisfy parameter inequalities (e.g., (4.2), (6.10), (8.7)).
  • b = 1 < b < (beta + sqrt(4 beta - 3 beta^2))/(4 beta), b close to 1
    Growth rate of frequencies lambda_{q+1} approximately lambda_q^b; needed to make the frequency separation and the Reynolds stress decay; the upper bound comes from the parameter inequality (8.7) for I2.
  • alpha = sufficiently small (alpha tends to 0+)
    Auxiliary exponent used throughout, e.g., in (2.4) and in (8.7); must be small enough for the interval for b to be nonempty and for estimates like l <= lambda_q^{-1} to hold.
  • s = s tends to 1/2+
    Sobolev exponent in Lemma 8.1; enters the parameter inequality (8.7) for the temperature Reynolds stress.
assumptions (6)
  • standard math Classical existence, uniqueness and estimates for the smooth Euler equations with smooth forcing (5.2)
    Invoked in Section 5.1 to define v_i via (5.2), citing [1, Chapter 7]; the estimates (5.3) rely on it.
  • standard math Standard elliptic Schauder estimates and the inverse-divergence operator R with div R f = f for zero-mean f
    Used in Sections 2, 5.2, 8; the identity (5.15) is from [13].
  • standard math Linear parabolic Schauder theory and the energy estimates (C.2), (C.3), and Lemma C.2 for the transport-diffusion equation
    Used in Section 7 to control theta_q and theta_{q+1}; Lemma C.2 is cited from [21, Lemma 3.7].
  • standard math The maximum principle for the transport-diffusion equation, giving ||theta||_0 <= ||theta0||_0
    Used in the proof of Theorem 1.1 to bound theta and in the Schauder estimate for p.
  • standard math Mikado flow existence (Lemma 6.1) with estimates (6.6)
    The perturbation is built from Mikado flows; taken from [12].
  • domain assumption The initial datum theta0 is a smooth function of x3 alone with zero mean
    Assumed in Theorem 1.1; required for the explicit starting case (Section 3) and for the zero-mean properties used in Remark 5.4.

how reviews work

0 comments
Cite this review

Pith. "Pith review of H\"{o}lder continuous weak solutions of the 3D Boussinesq equation with thermal diffusion." pith.science (2026). https://pith.science/paper/ICIOJ35F

@misc{pith2026250602927,
  author       = {Pith},
  title        = {Pith review of: H\"older continuous weak solutions of the 3D Boussinesq equation with thermal diffusion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ICIOJ35F}},
  note         = {Machine review of arXiv:2506.02927}
}
abstract

In this paper, we show the existence of H\"{o}lder continuous periodic weak solutions of the 3D Boussinesq equation with thermal diffusion, which apprroximate the Onsager's critical spatial regularity and satisfy the prescribed kinetic energy. More precisely, for any smooth $e(t):[0,T]\rightarrow \mathbb{R}_+$ and $\beta\in (0, \frac{1}{3})$, there exist $v\in C^{\beta}([0,T]\times {\mathbb{T} }^3)$ and $ \theta\in C_t^{1,\frac{\beta}{2}}C_x^{2,\beta}([0,T]\times {\mathbb{T} }^3)$ which solve (\ref{e:boussinesq equation}) in the sense of distribution and satisfy \begin{align} e(t)=\int_{{{\mathbb{T} }^3}}|v(t,x)|^2dx, \quad \forall t\in [0,T].\nonumber \end{align}

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages

  1. [5]

    Buckmaster, C

    T. Buckmaster, C. de Lellis, L. Sz\'ekelyhidi, Jr., and V. Vicol. Onsager's conjecture for admissible weak solutions. Comm. Pure Appl. Math. , 72(2):229--274, 2019

    work page 2019

  2. [21]

    T. Luo, T. Tao, and L. Zhang. H\"older continuous weak solutions of the 2 D B oussinesq equation with thermal diffusion. Sci. China Math. , 67(8):1777--1806, 2024

    work page 2024

  3. [1]

    Bahouri, J.-Y

    H. Bahouri, J.-Y. Chemin, and R. Danchin. Fourier analysis and nonlinear partial differential equations , volume 343 of Grundlehren der mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences] . Springer, Heidelberg, 2011

    work page 2011

  4. [2]

    Buckmaster

    T. Buckmaster. Onsager's conjecture almost everywhere in time. Comm.Math.Phys. , 333(3):1175--1198, 2015

    work page 2015

  5. [3]

    Buckmaster, C

    T. Buckmaster, C. De Lellis, P. Isett, and L. Sz\'ekelyhidi, Jr. Anomalous dissipation for 1/5 - H \"older E uler flows. Ann. of Math. (2) , 182(1):127--172, 2015

    work page 2015

  6. [4]

    Buckmaster, C

    T. Buckmaster, C. De Lellis, and L. Sz\'ekelyhidi, Jr. Dissipative E uler flows with O nsager-critical spatial regularity. Comm. Pure Appl. Math. , 69(9):1613--1670, 2016

    work page 2016

  7. [6]

    Buckmaster, S

    T. Buckmaster, S. Shkoller, and V. Vicol. Nonuniqueness of weak solutions to the SQG equation. Comm. Pure Appl. Math. , 72(9):1809--1874, 2019

    work page 2019

  8. [7]

    Buckmaster and V

    T. Buckmaster and V. Vicol. Nonuniqueness of weak solutions to the N avier- S tokes equation. Ann. of Math. (2) , 189(1):101--144, 2019

    work page 2019

Show all 26 references
  1. [8]

    Cheskidov, P

    A. Cheskidov, P. Constantin, S. Friedlander, and R. Shvydkoy. Energy conservation and O nsager's conjecture for the E uler equations. Nonlinearity , 21(6):1233--1252, 2008

    2008

  2. [9]

    Choffrut

    A. Choffrut. h -principles for the incompressible E uler equations. Arch. Ration. Mech. Anal. , 210(1):133--163, 2013

    2013

  3. [10]

    Constantin, W

    P. Constantin, W. E, and E. S. Titi. Onsager's conjecture on the energy conservation for solutions of E uler's equation. Comm. Math. Phys. , 165(1):207--209, 1994

    1994

  4. [11]

    S. Daneri. Cauchy problem for dissipative H \"older solutions to the incompressible E uler equations. Comm. Math. Phys. , 329(2):745--786, 2014

    2014

  5. [12]

    Daneri and L

    S. Daneri and L. Sz\'ekelyhidi, Jr. Non-uniqueness and h-principle for H \"older-continuous weak solutions of the E uler equations. Arch. Ration. Mech. Anal. , 224(2):471--514, 2017

    2017

  6. [13]

    De Lellis and L

    C. De Lellis and L. Sz\'ekelyhidi, Jr. Dissipative continuous E uler flows. Invent. Math. , 193(2):377--407, 2013

    2013

  7. [14]

    De Lellis and L

    C. De Lellis and L. Sz\'ekelyhidi, Jr. Dissipative E uler flows and O nsager's conjecture. J. Eur. Math. Soc. (JEMS) , 16(7):1467--1505, 2014

    2014

  8. [15]

    Duchon and R

    J. Duchon and R. Robert. Inertial energy dissipation for weak solutions of incompressible E uler and N avier- S tokes equations. Nonlinearity , 13(1):249--255, 2000

    2000

  9. [16]

    P. Isett. H\"older continuous E uler flows in three dimensions with compact support in time , volume 196 of Annals of Mathematics Studies . Princeton University Press, Princeton, NJ, 2017

    2017

  10. [17]

    P. Isett. A proof of O nsager's conjecture. Ann. of Math. (2) , 188(3):871--963, 2018

    2018

  11. [18]

    Isett and S.-J

    P. Isett and S.-J. Oh. On nonperiodic E uler flows with H \"older regularity. Arch. Ration. Mech. Anal. , 221(2):725--804, 2016

    2016

  12. [19]

    Isett and V

    P. Isett and V. Vicol. H\"older continuous solutions of active scalar equations. Ann. PDE , 1(1):Art. 2, 77, 2015

    2015

  13. [20]

    T. Luo, T. Tao, and L. Zhang. Finite energy weak solutions of 2 D B oussinesq equations with diffusive temperature. Discrete Contin. Dyn. Syst. , 40(6):3737--3765, 2020

    2020

  14. [22]

    A. Majda. Introduction to PDE s and waves for the atmosphere and ocean , volume 9 of Courant Lecture Notes in Mathematics . New York University, Courant Institute of Mathematical Sciences, New York; American Mathematical Society, Providence, RI, 2003

    2003

  15. [23]

    C. Miao, Y. Nie, and W. Ye. On O nsager's type conjecture for the inviscid B oussinesq equations. J. Funct. Anal. , 287(7):Paper No. 110527, 52, 2024

    2024

  16. [24]

    M. Novack. Nonuniqueness of weak solutions to the 3 dimensional quasi-geostrophic equations. SIAM J. Math. Anal. , 52(4):3301--3349, 2020

    2020

  17. [25]

    Tao and L

    T. Tao and L. Zhang. H\"older continuous solutions of B oussinesq equations. Acta Math. Sci. Ser. B (Engl. Ed.) , 38(5):1591--1616, 2018

    2018

  18. [26]

    Xu and Z

    S. Xu and Z. Tan. H\"older continuous solutions of B oussinesq equations with O nsager-critical spatial regularity. Calc. Var. Partial Differential Equations , 64(1):Paper No. 7, 44, 2025

    2025

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.