On the two-dimensional Navier-Stokes equations with horizontal viscosity

Chongsheng Cao; Yanqiu Guo

REVIEW 2 major objections 6 minor 11 references

On the two-dimensional Navier-Stokes equations with horizontal viscosity

T0 review · 2 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper proves that the 2D Navier-Stokes equations with horizontal viscosity only have a unique global weak solution when the initial velocity lies in $L^2$ and its vertical derivative $\partial_y u_0$ lies in $L^2$, lowering the…

desk verdict Interesting regularity improvement with strong formal estimates, but the existence proof's approximation sequence is not divergence-free and the function space is too restrictive as stated. read the letter →

arxiv 2507.02775 v1 pith:LGXW64F6 submitted 2025-07-03 math.AP

Chongsheng Cao , Yanqiu Guo This is my paper · ORCID

classification math.AP MSC 35Q3076D0535B6576D03

keywords Navier-Stokesequationshorizontalviscosityanisotropicglobalweaksolutionschannelflowwell-posednessaprioriestimatesshearstability

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 considers the 2D Navier-Stokes equations with viscosity acting only in the horizontal direction, in a channel that is periodic horizontally and bounded by impenetrable walls above and below. Its main result is that these equations have a unique global weak solution whenever the initial velocity lies in $L^2$ and its vertical derivative $\partial_y u_0$ also lies in $L^2$; no differentiability of the vertical velocity component is required beyond square-integrability. This lowers the regularity threshold of an earlier global well-posedness result, which required the vertical derivative of both velocity components to be in $L^2$. The paper also proves a uniform-in-time $H^2$ bound under stronger assumptions, exponential decay of oscillatory vorticity around a shear flow, and convergence to a steady shear profile in the unforced case. The novelty relative to the earlier result is that only one vertical derivative of the initial velocity is needed.

What carries the argument

The load-bearing mechanism is an a priori estimate for the vertical derivative of the horizontal velocity, $\|u_y\|_2$. The authors differentiate the first component of the momentum equation in $y$, introduce the pressure Poisson equation with boundary condition $p_y|_{y=0}=p_y|_{y=1}=0$, and bound the pressure term by solving $-\Delta\phi=u_y$ with zero Dirichlet boundary conditions on $\phi$. The resulting differential inequality (3.28) for $\|u_y\|_2^2$ has a Grönwall factor built from quantities controlled by the basic energy estimate, so it closes without assuming $\partial_y v_0\in L^2$. Trilinear estimates based on Agmon-type inequalities and a Poincaré inequality for $v$ (using $v|_{y=0}=0$) keep every term at this lower regularity. Uniqueness is then obtained by subtracting two weak solutions and using the same bounds in a Grönwall inequality.

What would settle it

If one can exhibit finite-mode analytic initial data and forcing for which the truncated system (3.33) forms a singularity in finite time, or otherwise fails to have a global analytic solution, then the approximation argument in Section 3.1.3 collapses, and the a priori estimates alone would not produce a convergent sequence of approximate solutions.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for initial velocity $(u_0,v_0)\in H$ with $\partial_y u_0\in L^2(\Omega)$ and forcing $f\in L^1(0,T;H)$ with $\partial_y f_1\in L^2(0,T;L^2(\Omega))$, equation (1.1) has a unique global weak solution for every $T>0$, with regularity $u_y\in L^\infty(0,T;L^2)$ and $u_x, v_x, u_{xy}\in L^2(0,T;L^2)$. The paper's secondary claims are a uniform bound on $\|u(t)\|_{H^2}$ for all time under $u_0\in H\cap H^2$ and an integrability condition on $f$ with $f_1=0$ (Theorem 1.2), exponential decay of the oscillatory part of the vorticity perturbation around the shear flow $(ay,0)$ (Proposition 1.3), and convergence to a steady shear profile $(\phi(y),0)$ when the forcing vanishes (Proposition 1.4).

Load-bearing premise

The argument assumes, without proof or citation, that the finite-mode Galerkin system (3.33) with analytic initial data and analytic forcing has a global analytic solution on every finite time interval, justified only by the statement that it is 'more regular than 2D Euler equations'.

Editorial extensions

If this is right

The initial regularity needed for global well-posedness drops from $\partial_y u_0,\partial_y v_0\in L^2$ to just $\partial_y u_0\in L^2$ (with $u_0,v_0\in L^2$), so rougher data are admissible.
Because the weak solutions are unique in this larger regularity class, the solution map is single-valued on the newly admitted data.
Under $u_0\in H^2$ and the stated integrability assumptions on $f$ with $f_1=0$, the $H^2$ norm of the velocity stays bounded by a constant depending only on $\|u_0\|_{H^2}$ and $C_0$ for all time.
Small perturbations of the shear flow $(ay,0)$ have oscillatory vorticity decaying exponentially when the forcing is curl-free and the initial vorticity is small, so the shear flow is stable in that sense.
With zero forcing, every global solution converges in $H$ to a steady shear profile $(\phi(y),0)$, where $\phi$ is the limit of the horizontal mean of $u$.

Reading between the lines

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

The one-sided derivative mechanism may transfer to other anisotropic dissipation geometries, because the key opens with a bound on $\|u_y\|_2$ plus a pressure-Poisson auxiliary problem rather than requiring symmetric differentiability of both velocity components.
The unproved assertion of global analytic well-posedness for the truncated Galerkin system (3.33) is the natural place to probe the proof; a direct existence argument for finite-mode analytic solutions would make the approximation step self-contained.
Because uniqueness holds below the prior regularity threshold, continuous dependence on initial data and forcing in the enlarged class is a plausible next property, though the paper does not establish a modulus of continuity.
The exponential decay of the oscillatory vorticity near shear flows suggests that damping of the horizontal-mean part is what drives the stability; quantifying the smallness condition (5.10) would be a direct extension.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Two things to know about this paper. First, the claimed improvement over Liang–Zhang–Zhu is real: global well-posedness under only ∂_y u_0 in L^2 (instead of both ∂_y u_0 and ∂_y v_0) is a meaningful step, and the uniform H^2 bound and shear-stability results are genuinely new. Second, the proof as written has a load-bearing gap in the approximation argument, and it is a bit different from the one the stress-test flagged.

The a priori estimates themselves are the strongest part. The pressure handling via the auxiliary function φ is clever, the calculus is detailed, and the resulting bounds are internally consistent. The uniqueness proof also looks sound given the stated regularity.

The trouble is Section 3.1.3. The authors truncate the Fourier series of u_0 and v_0 separately and claim the truncated pair approximates u_0. But that pair is not divergence-free in general, so it does not belong to the space H defined in (1.3). The divergence-free condition couples u and v through ∂_x u + ∂_y v = 0; truncating the two series independently breaks the constraint. And the space H itself is not the full L^2 divergence-free space for the channel: with u expanded in e^{2πi k_2 y} and v in sin(π k_2 y), the condition forces the odd k_2 modes of v to vanish, so the theorem only covers a proper subspace, despite the abstract's claim of arbitrary L^2 data.

The stress-test note is right that the asserted global analytic solvability of (3.33) is unproved, but the non-divergence-free initial data is a prior problem. You cannot even start the incompressible equation (3.33) with ∇·u_m = 0 from such initial data. This is repairable—use a proper Galerkin projection onto a divergence-free basis, or work with the stream function—but it is not a minor typo.

Minor issue: Theorem 1.2 assumes f_1 = 0 and ∂_{yy} f_1 ∈ L^2, which is redundant; likely a typo for ∂_y f_2 or similar.

Overall: if the approximation argument is fixed, the main theorem is likely true and well within the subfield's standards. As printed, the existence proof is not complete. I would send it to a referee, mainly to have them pressure the authors to rework Section 3.1.3 and restate H properly. Not something to cite until that happens.

Referee Report

2 major / 6 minor

Summary. The paper studies the two-dimensional Navier-Stokes equations with horizontal viscosity only in the periodic-in-x channel Ω=T×[0,1] with impermeable walls. The main result (Theorem 1.1) asserts global well-posedness of weak solutions under the low-regularity assumptions u0,v0∈L2 and ∂y u0∈L2, with forcing f∈L1(0,T;H) and ∂y f1∈L2(0,T;L2). The proof is based on a priori estimates for ∥u∥2 and ∥uy∥2 together with a Galerkin approximation. Further results give a uniform H2 bound under an integrability condition on the forcing (Theorem 1.2), exponential decay of the fluctuation part of the vorticity around a shear flow (Proposition 1.3), and convergence to a shear steady state for unforced flows (Proposition 1.4).

Significance. If Theorem 1.1 is correct, it improves the previous global well-posedness threshold for this anisotropic viscosity model, which required both ∂y u0 and ∂y v0 in L2. The a priori estimates are detailed and internally consistent, and the paper provides a concrete new regularity class for the problem. The secondary results (H2 bound, shear stability, asymptotic convergence) are natural complements and appear to follow from the developed estimates. However, the gap in the approximation argument described below currently prevents the main theorem from being fully established.

major comments (2)

[§3.1.3] The existence of global solutions to the truncated Galerkin system (3.33) is asserted without proof or citation, on the grounds that the system is 'more regular than 2D Euler equations.' This does not constitute a proof: (3.33) is a different equation with different boundary conditions, and no theorem is stated that covers it. Since the entire existence proof of Theorem 1.1 relies on the availability of the approximating sequence u_m on arbitrary time intervals, the authors must either supply a standard ODE-based local existence argument (the system is finite-dimensional after projection) together with the energy bound (3.34) to rule out blow-up, or cite a suitable well-posedness theorem.
[§3.1.2–§3.1.3] The key estimate (3.28) is derived using the pressure boundary condition (3.8) and the pressure Poisson equation (3.7). For the Galerkin projection (3.33), the pressure is not part of the projected system; any pressure reconstructed from the residual need not satisfy py=0 at y=0,1. The statement in §3.1.3 that 'all the a priori estimates conducted in subsections 3.1.1 and 3.1.2 are valid for um' is therefore not justified as written. The authors should either prove (3.28) for the projected system without using the pressure boundary condition, or choose an approximating scheme (e.g., with a compatible pressure or a different regularization) for which (3.8) holds.

minor comments (6)

[Theorem 1.2] The assumption f1=0 makes the term ∥∂yy f1∥2 in the integrability condition identically zero; this is likely a typo for a condition on ∂y f2 or ∂xy f2. Please state the intended hypothesis.
[§3.1.3] The displayed line before (3.39), '∂xum = ∂yvm', should read '∂y vm = −∂x um' (sign).
[§3.1.3] The notation ∂xum ∈ L2(0,T;H) in (3.38) is ambiguous; it should be made explicit that it denotes the pair (∂xum, ∂xvm).
[Section 5] The notations fω∗ and ω∗ in (5.2)–(5.8) are confusing; presumably \widetilde{\omega^*} and \overline{\omega^*} are intended. Please use consistent notation for the mean and fluctuation parts.
[Section 4.2] In (4.15), the term involving ∥∂yy f1∥2 is redundant under the hypothesis f1=0 of Theorem 1.2; see the earlier comment on the theorem statement.
[Throughout] The paper contains numerous typographical errors (title, 'Poinc´ are', 'Gronwall', etc.); a careful proofreading is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; Section 3.1.3 has an unsupported existence assertion, but it is a gap, not a circular step.

full rationale

The derivation chain in Theorem 1.1 is not circular. The a priori estimates (3.1)-(3.3) and (3.12)-(3.30) are obtained from the PDE by standard inner-product estimates and the stated boundary conditions; the pressure term is rewritten through the Poisson problem (3.7)-(3.11) without importing the theorem being proved. The uniqueness argument in Section 3.2 is a Gronwall estimate on the difference of two solutions using only the previously established bounds. The later results (Theorem 1.2, Propositions 1.3 and 1.4) are corollaries of the same estimates and vorticity identities, not re-statements of the assumptions. The only notable logical gap is Section 3.1.3's assertion that the truncated system (3.33) has a global analytic solution because it is 'more regular than 2D Euler equations'; this is an unsupported existence premise (and the pressure boundary condition for the truncated system is not justified), but it is not circular: it does not assume Theorem 1.1, is not a fitted parameter, and is not justified by a self-citation. The paper contains no self-citations at all, and the cited external results are used as benchmarks or standard background facts, not as load-bearing substitutes for the proof. Hence the circularity score is 0.

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

The central claims rest on standard inequalities and the specific boundary structure of the channel. No free parameters or invented entities are introduced. The one unproved ingredient is the analytic solvability of the truncated Galerkin system in §3.1.3, which is standard but asserted without proof. The pressure boundary condition from (3.8) is another load-bearing modeling assumption inherited from the forcing and boundary conditions.

assumptions (3)

standard math Global analytic solvability of the finitely truncated system (3.33) for analytic initial data and forcing
Invoked in §3.1.3 without proof or citation. Standard for finite-dimensional ODE systems with polynomial nonlinearities and for analytic 2D Euler, but not demonstrated in the paper.
domain assumption Pressure boundary condition ∂_y p = 0 at y=0,1 derived from (3.5) and the assumptions v=0, f2=0 on the walls
Used to justify the pressure estimates in §3.1.2; relies on the forcing boundary condition f2|y=0 = f2|y=1 = 0.
standard math Poincaré, Agmon and Ladyzhenskaya inequalities on the channel
Used throughout; standard on bounded domains with the stated boundary conditions.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On the two-dimensional Navier-Stokes equations with horizontal viscosity." pith.science (2026). https://pith.science/paper/LGXW64F6

@misc{pith2026250702775,
  author       = {Pith},
  title        = {Pith review of: On the two-dimensional Navier-Stokes equations with horizontal viscosity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LGXW64F6}},
  note         = {Machine review of arXiv:2507.02775}
}

abstract

This paper is concerned with a 2D channel flow that is periodic horizontally but bounded above and below by hard walls. We assume the presence of horizontal viscosity only. We study the well-posedness, large-time behavior, and stability of solutions. For global well-posedness, we aim to assume less differentiability on initial velocity $(u_0, v_0)$: in particular, we assume $u_0,v_0\in L^2(\Omega)$ and $\partial_y u_0 \in L^2(\Omega)$.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

11 extracted references · 8 canonical work pages

[1]

Albritton, E

D. Albritton, E. Bru´ e, and M. Colombo,Non-uniqueness of Leray solutions of the forced Navier-Stokes equations, Ann. of Math. (2) 196 (2022), 415–455

work page 2022
[2]

Albritton and M

D. Albritton and M. Colombo, Non-uniqueness of Leray solutions to the hypodissipative Navier-Stokes equations in two dimensions , Comm. Math. Phys. 402 (2023), 429–446

work page 2023
[3]

C´ ordoba, L

D. C´ ordoba, L. Martinez-Zoroa, and F. Zheng,Finite time blow-up for the hypodissipative Navier Stokes equations with a force in L1 t C 1,ϵ x ∩ L∞ t L2 x, arXiv:2407.06776 [math.AP]

arXiv
[4]

B. Dong, J. Wu, X. Xu, and N. Zhu, Stability and exponential decay for the 2D anisotropic Navier-Stokes equations with horizontal dissipation , J. Math. Fluid Mech. 23 (2021), Paper No. 100, 11 pp

work page 2021
[5]

Elgindi, Finite-time singularity formation for C 1,α solutions to the incompressible Euler equations on R3, Ann

T. Elgindi, Finite-time singularity formation for C 1,α solutions to the incompressible Euler equations on R3, Ann. of Math. (2) 194 (2021), 647–727

work page 2021
[6]

Grafakos, Classical Fourier analysis, Second edition, Graduate Texts in Mathematics 249, Springer, New York, 2008

L. Grafakos, Classical Fourier analysis, Second edition, Graduate Texts in Mathematics 249, Springer, New York, 2008. 2D NSE WITH HORIZONTAL VISCOSITY 21

work page 2008
[7]

Kiselev and V

A. Kiselev and V. ˇSver´ ak,Small scale creation for solutions of the incompressible two-dimensional Euler equation, Ann. of Math. (2) 180 (2014), 1205–1220

work page 2014
[8]

Liang, P

S. Liang, P. Zhang, and R. Zhu, Deterministic and stochastic 2D Navier-Stokes equations with anisotropic viscosity, J. Differential Equations 275 (2021), 473–508

work page 2021

Show all 11 references

[9]

V. I. Yudovich, Non-stationary flows of an ideal incompressible fluid, ˇZ. Vyˇ cisl. Mat i Mat. Fiz. 3 (1963), 1032–1066

1963
[10]

Vishik, Instability and non-uniqueness in the Cauchy problem for the Euler equations of an ideal incompressible fluid

M. Vishik, Instability and non-uniqueness in the Cauchy problem for the Euler equations of an ideal incompressible fluid. Part I , arXiv:1805.09426 [math.AP]

arXiv
[11]

Vishik, Instability and non-uniqueness in the Cauchy problem for the Euler equations of an ideal incompressible fluid

M. Vishik, Instability and non-uniqueness in the Cauchy problem for the Euler equations of an ideal incompressible fluid. Part II , arXiv:1805.09440 [math.AP]. Department of Mathematics & Statistics, Florida International University, Miami, Florida 33199, USA Email address : c...

arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Albritton, E

D. Albritton, E. Bru´ e, and M. Colombo,Non-uniqueness of Leray solutions of the forced Navier-Stokes equations, Ann. of Math. (2) 196 (2022), 415–455

work page 2022

[2] [2]

Albritton and M

D. Albritton and M. Colombo, Non-uniqueness of Leray solutions to the hypodissipative Navier-Stokes equations in two dimensions , Comm. Math. Phys. 402 (2023), 429–446

work page 2023

[3] [3]

C´ ordoba, L

D. C´ ordoba, L. Martinez-Zoroa, and F. Zheng,Finite time blow-up for the hypodissipative Navier Stokes equations with a force in L1 t C 1,ϵ x ∩ L∞ t L2 x, arXiv:2407.06776 [math.AP]

arXiv

[4] [4]

B. Dong, J. Wu, X. Xu, and N. Zhu, Stability and exponential decay for the 2D anisotropic Navier-Stokes equations with horizontal dissipation , J. Math. Fluid Mech. 23 (2021), Paper No. 100, 11 pp

work page 2021

[5] [5]

Elgindi, Finite-time singularity formation for C 1,α solutions to the incompressible Euler equations on R3, Ann

T. Elgindi, Finite-time singularity formation for C 1,α solutions to the incompressible Euler equations on R3, Ann. of Math. (2) 194 (2021), 647–727

work page 2021

[6] [6]

Grafakos, Classical Fourier analysis, Second edition, Graduate Texts in Mathematics 249, Springer, New York, 2008

L. Grafakos, Classical Fourier analysis, Second edition, Graduate Texts in Mathematics 249, Springer, New York, 2008. 2D NSE WITH HORIZONTAL VISCOSITY 21

work page 2008

[7] [7]

Kiselev and V

A. Kiselev and V. ˇSver´ ak,Small scale creation for solutions of the incompressible two-dimensional Euler equation, Ann. of Math. (2) 180 (2014), 1205–1220

work page 2014

[8] [8]

Liang, P

S. Liang, P. Zhang, and R. Zhu, Deterministic and stochastic 2D Navier-Stokes equations with anisotropic viscosity, J. Differential Equations 275 (2021), 473–508

work page 2021