{"id":"f7818f77-5419-45b1-94bf-0d6e24253f9d","arxiv_id":"2608.04621","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A new theorem establishes local-in-time existence, uniqueness, and spatial decay estimates for Navier-Stokes solutions in a half-space with initial data in a class of weighted Lebesgue spaces.","lead":"This paper proves that the Navier-Stokes equations in a half-space admit a smooth, unique solution for a short time when the starting flow belongs to a naturally weighted function space. The result matters because it provides rigorous spatial decay rates for the velocity field and extends a known technique from the whole-space setting to the physically relevant half-space problem.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4 (Eq. 33) and Lemma 9 (Eq. 69) are false for weights with centers away from the origin; the metric (7) is incompatible with the general weight (2).","rationale":"The paper's central theorem claims local well-posedness for the Navier-Stokes IBVP in the half-space with the general weight (2), i.e., a product of powers of distances from finitely many arbitrary centers. The entire proof is organized around the metric (7), which uses the origin-centered factor |x|^alpha, and around pointwise estimates (33) and (69) that give t^{n/(2p)}|x|^alpha decay. These estimates cannot hold for a weight with a nonzero center: the weight vanishes at that center, so admissible initial data may have a local singularity there, and the heat/Stokes evolution at that point decays only like t^{-beta/2} for beta<1. Choosing beta with 1-alpha < beta < 1 makes the left side of (33) diverge as t->0 while K_g(t,rho) remains bounded. This is not a technical gap in a peripheral estimate; it is an explicit counterexample to the key linear estimate used to start the contraction argument in Lemma 15 and to control the metric in Lemma 17. Consequently Theorem 1 as stated for the general weight (2) is not supported. The theorem may survive if the weight is restricted to a single center at the origin (the radial setting of [20]), or if the metric is reformulated using w(x) instead of |x|^alpha, but either change alters the advertised statement. The reader's weakest assumption concerned the external kernel estimates (38) and (52); while that is also a legitimate concern about documentation of cited bounds, the false estimate (33)/(69) is more load-bearing because it is an internal inconsistency in the submitted proof. Therefore the verdict should move from CONDITIONAL to REJECT, with the caveat that a restricted version may be recoverable.","tokens_in":28354,"tokens_out":27643,"duration_ms":333725,"concrete_test":"Verify Lemma 4, Eq. (33), analytically or numerically. Set n=3, p=6, alpha=1/2, weight w(y)=|y-(1,0,0)|^{1/2}, and g(y)=|y-(1,0,0)|^{-0.9} on the unit ball centered at (1,0,0), zero elsewhere. Compute H(t)*g at x=(1,0,0) for t=10^{-k}, k=3,...,8. If t^{1/4}|H*g(t,(1,0,0))| grows roughly like t^{-0.2} while K_g(t,1) remains bounded, then (33) is false. Equivalently, check whether t^{1/4}|H*g(t,(1,0,0))| remains bounded as t->0; it will diverge, confirming the failure.","verdict_should_be":"REJECT","load_bearing_attack":"The proof of Theorem 1 rests on the pointwise estimates (33) and (69), which assert global decay of the Stokes/heat evolution as t^{n/(2p)}|x|^alpha |u(t,x)| <= K_g(t,rho). For the general weight (2), w(x)=prod|x-bar{x}_j|^{alpha_j} with nonzero centers, this is internally inconsistent. Since w vanishes at each center bar{x}_j, the norm L^p_w does not control local size near bar{x}_j, and data with a singularity there are admissible. Concretely, take n=3, p=6 (alpha=1/2), weight w(y)=|y-bar{e}_1|^{1/2}, and g(y)=|y-bar{e}_1|^{-beta} on B(bar{e}_1,1) with beta=0.9. Then g in L^p_w, but (H(t)*g)(t,bar{e}_1) is comparable to t^{-beta/2}=t^{-0.45}, so the left side of (33) at x=bar{e}_1 behaves like t^{n/(2p)}|x|^alpha |H*g| ~ t^{0.25-0.45}=t^{-0.2}, which diverges as t->0. Meanwhile K_g(t,rho) stays bounded: the local term c_0|g|_{w,p,rho} is finite and the exponential term c_1 e^{-rho^2/(8t)}|g|_{w,p} vanishes for fixed rho. Thus (33) is false, and Lemma 9 inherits the failure. The displayed proof of Lemma 9 drops the heat-kernel term when deriving the |x|^{-alpha} bound, and no argument using the actual weight w(x) is supplied. Lemmas 15-17 and the main theorem use (69) at every stage, so the advertised result for arbitrary finite centers is not established. This is an internal false estimate, not merely a reliance on quoted kernel estimates (38) and (52).","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the initial-boundary value problem for the Navier-Stokes equations in the half-space R^n_+ with initial data in a weighted Lebesgue space L^p_w(R^n_+), where the weight is a product of powers of distances from finitely many fixed centers (Eq. (2)). The main results, Theorems 1 and 2, claim local existence for large divergence-free data and global existence for small data, together with L^q estimates, pointwise spatial decay of the form t^{n/(2p)}|x|^\\alpha|u(t,x)| bounded by a quantity K(t,\\rho), and uniqueness in the same class. The proof is built on a successive-approximation scheme in the metric (7), using the Stokes semigroup estimates developed in the authors' previous works and pointwise kernel estimates quoted from Solonnikov and Crispo-Maremonti.","tokens_in":28740,"tokens_out":8679,"duration_ms":110846,"significance":"If valid, the result would meaningfully extend the authors' earlier Cauchy-problem theory to the half-space for weights with multiple singular centers. The paper is well organized, the nonlinear-term decomposition in Section 4.2 is clearly structured, and the duality-based uniqueness argument follows a plausible known pattern. However, a load-bearing pointwise estimate is false for the general weight (2), and the advertised quantitative decay estimate is therefore not established. The claimed generalization to weights with arbitrary centers cannot be accepted on the present proof.","major_comments":[{"comment":"The pointwise estimate (33) is false as stated for weights with centers away from the origin. Take n=3, p=6 (so alpha=1/2), w(y)=|y-\\bar e_1|^{1/2}, and g(y)=|y-\\bar e_1|^{-0.9} on B(\\bar e_1,1), extended by zero. Then g belongs to L^p_w(R^3), and for fixed rho the right-hand side K_g(t,rho) of (5) is bounded uniformly in t. However, (H*g)(t,\\bar e_1) is comparable to t^{-0.45} as t tends to 0, so t^{n/(2p)}|\\bar e_1|^\\alpha (H*g)(t,\\bar e_1) behaves like t^{-0.2} and diverges, contradicting (33). The omitted proof in [20] is for the radial weight |y|^\\alpha, for which the singular center coincides with the origin; for the general weight (2) there is no relation between |x| and w(x), so local singularities at nonzero centers are not controlled by the right-hand side of (33).","section":"Section 3, Lemma 4, Eq. (33)"},{"comment":"Lemma 9 inherits the same defect. The proof of (69) invokes the false estimate (33) for the odd heat-extension term, and the displayed argument does not perform a cancellation with the G^* term. With the same data as above, placed in the half-space with center \\bar e_1=(1,0,1), say, the first term in the expression for phi(t,x) behaves like t^{-0.45} at x=\\bar e_1, while K_g(t,rho) remains bounded. Consequently (69) fails. This is not a cosmetic gap: Lemmas 10, 15, 16, and 17 and the proof of Theorem 1 use (69) to control the |x|^\\alpha component of the metric (7), to obtain (10), and to prove the limit property (11). The proof of the main theorem therefore does not go through for the class of weights announced in (2), and the asserted bound (10) is not supported for admissible data with integrable singularities at nonzero centers.","section":"Section 4.1, Lemma 9, Eq. (69)"}],"minor_comments":[{"comment":"The displayed assumption in Lemma 13 is incomplete: the two suprema are written without the condition that they are finite, so the statement as displayed does not parse.","section":"Section 4.2, Lemma 13"},{"comment":"The symbol u_0 is used both for the initial datum and for the linear solution G[u_0]; this overloading makes the iteration estimates in Lemma 15 harder to follow.","section":"Section 5, after Eq. (92)"},{"comment":"There are several typographical slips, including 'looking for' in the Introduction, 'corrisponding' in Lemma 17, and 'easly' in Lemma 14.","section":"Throughout"},{"comment":"The reference list contains a formatting artifact in entry [2], where 'J., 41, pp. 1027-1076 (1992)' is attached after the title.","section":"References"}],"recommendation":"reject","confidential_remarks":"The flaw is not a missing detail that a revision could patch locally: Eq. (33) is false for the very class of weights introduced in (2), and Eq. (69) is used at every stage of the iteration. Since Lemma 4 is stated without proof as 'analogous' to [20], the transfer from the radial weight to the general multi-center weight appears to be the source of the problem. The authors should verify whether the companion paper [3] contains the same pointwise estimate in the general-weight setting before resubmitting."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper aims to complete a program: extending the weighted Lebesgue space Navier-Stokes theory to the half-space IBVP, with weights that are products of powers of distances to several points. The iteration scheme and the duality uniqueness argument follow known patterns, and the paper is clearly written. But there is a load-bearing false estimate, and I do not think the main theorem is proven as stated.\n\nThe problem is Lemma 4, equation (33), and its half-space counterpart Lemma 9, equation (69). They claim that for any g in L^p_w, t^{n/(2p)}|x|^alpha |H*g(t,x)| <= K_g(t,rho). This fails for nonzero weight centers. Take n=3, p=6 (alpha=1/2), weight w(y)=|y-x0|^{1/2} with x0 not equal to 0, and g(y)=|y-x0|^{-0.9} on a small ball around x0. Then g is in L^p_w, but (H*g)(t,x0) ~ t^{-0.45}. So the left side of (33) at x=x0 behaves like t^{1/4-0.45}=t^{-0.2}, which blows up as t->0. Meanwhile K_g(t,rho) stays bounded: the local term is finite and the exponential term vanishes. So the estimate is simply false for the generality the authors claim. Lemma 4's proof is omitted, and Lemma 9's proof invokes (33) directly. Lemmas 14-17 and the proof of Theorem 1 all lean on (69). The metric (7), which uses |x|^alpha, is incompatible with weights whose zeros are away from the origin.\n\nThe reader's report flagged the reliance on quoted kernel estimates (38) and (52) from Solonnikov and Crispo-Maremonti. That is a separate gap, but it is fixable. The false pointwise estimate is more serious; no amount of restating the quoted bounds will repair it.\n\nWhat the paper does well: the extension from the Stokes IBVP to the nonlinear Navier-Stokes problem is a natural next step, the quantitative estimates are spelled out, and the authors are honest about the qualitative existence time. If the class of weights were restricted (say, to centers at the origin), the argument would likely go through. But as written, the advertised result for arbitrary finite centers is not proven.\n\nI would send this to referees rather than desk-reject: the problem is worthwhile, the flaw is specific and testable, and a revision that fixes the metric or restricts the weights could produce a correct paper. But my own recommendation on the present version is reject; the central estimate needs to be reworked.\n\nBest","headline":"The paper extends a worthwhile program to the half-space, but a central pointwise estimate is false for general weights with nonzero centers, so the main theorem is not established.","tokens_in":29317,"tokens_out":6516,"would_cite":false,"duration_ms":66163,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q30","76D03","76D05"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper establishes local existence, uniqueness, and quantitative decay for half-space Navier-Stokes solutions with divergence-free initial data in a weighted Lebesgue space L^p_w(R^n_+), p>n, and global existence for suitably small data.","keywords":["Navier-Stokes equations","half-space","initial-boundary value problem","weighted Lebesgue spaces","Stokes semigroup","scale-invariant weights","mild solutions","uniqueness"],"falsifier":"Check the two quoted kernel estimates directly for a weight whose singular point approaches the boundary: if the constant in (38) or (52) fails to stay bounded as the singular point approaches {x_n=0}, or if (52) fails for n=3 when |l'|=1, then the L^q and pointwise Lemmas 11-14 would not close and Theorem 1 would not follow as stated.","tokens_in":28111,"feed_emoji":"🌊","tokens_out":5990,"duration_ms":70984,"temperature":0.7,"pith_summary":"This paper aims to show that the Navier-Stokes initial-boundary value problem in the upper half-space is locally well posed for initial data lying in a weighted Lebesgue space L^p_w(R^n_+), p>n, where the weight is a product of powers of distances to finitely many fixed interior points. The authors construct a smooth solution via integral representation formulas for velocity and pressure, prove quantitative L^q and pointwise decay estimates on a time interval determined by the data, and show the solution converges to the prescribed initial datum at t=0. They also prove uniqueness among solutions in this class, and global existence when the weighted norm is suitably small. If correct, this gives a half-space, boundary-value counterpart of the weighted Cauchy-problem theory and extends the Stokes semigroup analysis to the nonlinear Navier-Stokes system.","feed_headline":"Half-space Navier-Stokes flows from singular weighted data","feed_subtitle":"Divergence-free data in weighted L^p, p>n, give smooth local solutions with spatial decay; small data last forever.","key_machinery":"The argument is carried by the half-space Green function G, the pressure kernels Q and K, and the scale-invariant metric (7) that measures weighted |x|^$\\alpha$ decay, L^infty decay, L^q decay, and gradient decay. Solutions are built by successive approximation, u_m = G[u0] - S[u_{m-1}·nabla u_{m-1}], and the central estimates Lemmas 11-14 turn the convective term into quadratic expressions in this metric, while weighted Stokes semigroup estimates from the authors' earlier work control the linear term. The fixed point gives the a priori bound (10), and the t tending to 0 limit is obtained through an absolute-continuity argument that yields only a qualitative description of T(u0).","core_discovery":"The central claim is Theorem 1: for divergence-free u0 in L^p_w(R^n_+), p>n, there exists a time T(u0)>0 and a smooth solution (u,π) of the Navier-Stokes IBVP, expressed by the representation formulas (8)-(9), satisfying the four-term estimate (10) with constants controlled by K(t,rho) and the weighted norm of u0, the limit (11) in the chosen metric, the dual-space convergence (12), and the pressure estimate (13). Theorem 2 asserts uniqueness of this solution in the class detected by Theorem 1. For suitably small weighted norm, the same results hold for all positive times.","pith_inferences":["The same iteration could plausibly be extended to weights whose singular points approach or reach the boundary, but only if the quoted pointwise kernel estimates (38) and (52) remain valid with constants independent of the distance from the singular points to {x_n=0}; checking that is a natural next step.","Because the proof does not give a quantitative lower bound for T(u0), a testable project is to track the constants in the algebraic fixed-point lemma and express T in terms of the local weighted mass K(t,rho) or a modulus of continuity of the absolute-continuity argument.","The metric-based iteration is not tied to a specific semigroup representation, so the same scheme may adapt to other parabolic initial-boundary value problems with similar boundary kernels, such as MHD or Boussinesq systems in a half-space."],"forward_implications":["Every divergence-free initial datum in the weighted space produces a smooth Navier-Stokes flow on a short time interval, so this weighted Lebesgue space is an admissible initial-data class for the half-space problem.","The solution carries explicit spatial decay: for fixed t the estimate t^{n/(2p)}|x|^alpha |u(t,x)| is controlled by data-dependent quantities, so the velocity decays at least like |x|^{-(1-n/p)} away from the singular points.","When the weighted norm of the initial datum is suitably small, the same bounds hold for all t>0, giving global existence of smooth solutions in the small-data regime.","The pressure gradient belongs to L^r(eta,T;L^q) for r>1, q>n, with a bound quadratic in the same data-dependent quantity, so the pressure term inherits the regularity of the velocity construction.","Uniqueness holds within the class, meaning the solution constructed by the iteration is the only solution in that regularity class with the same initial datum."],"supporting_citations":[{"why":"Supplies the weighted Stokes semigroup estimates, Theorem 5 and Corollary 1, used to control the linear part of the iteration.","marker":"[3]"},{"why":"Provides the pointwise estimate (52) for the composite kernel K and the half-space Navier-Stokes theory that the later kernel bounds build on.","marker":"[2]"},{"why":"Establishes the pointwise estimates (38) for the kernels G* and Q and the Stokes representation theory on which the whole estimate chain rests.","marker":"[27]"},{"why":"Supplies the iterative scheme, the generalized Gronwall lemma, and the uniqueness strategy adapted here to the half-space.","marker":"[20]"},{"why":"Gives the existence and representation theory for the nonstationary Stokes and Navier-Stokes systems and the algebraic lemma used to close the iteration.","marker":"[25]"},{"why":"Provides the absolute-continuity property (P) used to obtain the t tending to 0 limit and the qualitative dependence of T(u0) on the data.","marker":"[1]"},{"why":"Supplies the duality argument at the core of the uniqueness proof in Theorem 2.","marker":"[6]"}],"fun_headline_variants":["Singular weights, smooth flows: half-space Navier-Stokes","Weighted L^p data: local smooth Navier-Stokes in half-space","Local well-posedness for Navier-Stokes in weighted half-space","Half-space Navier-Stokes: from singular data to smooth flows","Navier-Stokes IBVP: weighted data give smooth local solutions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the pointwise kernel estimates (38) and (52), quoted without proof from earlier work, hold for all data and weights used here; if those estimates require extra restrictions on the weight singularities or on initial-data decay, the main theorem would not follow as stated.","fun_headline_variants_meta":{"raw":{"variants":["Singular weights, smooth flows: half-space Navier-Stokes","Weighted L^p data: local smooth Navier-Stokes in half-space","Local well-posedness for Navier-Stokes in weighted half-space","Half-space Navier-Stokes: from singular data to smooth flows","Navier-Stokes IBVP: weighted data give smooth local solutions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000745,"raw_usage":{"total_tokens":3295,"prompt_tokens":890,"completion_tokens":2405,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":506,"completion_tokens_details":{"reasoning_tokens":2310}},"tokens_in":506,"tokens_out":2405,"duration_ms":18437,"temperature":1.0,"reasoning_tokens":2310,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:29:36.330735+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the two quoted kernel estimates directly for a weight whose singular point approaches the boundary: if the constant in (38) or (52) fails to stay bounded as the singular point approaches {x_n=0}, or if (52) fails for n=3 when |l'|=1, then the L^q and pointwise Lemmas 11-14 would not close and Theorem 1 would not follow as stated.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the weighted Stokes semigroup estimates, Theorem 5 and Corollary 1, used to control the linear part of the iteration."},{"cited_title":"Crispo and P","cited_arxiv_id":null,"evidence_quote":"Provides the pointwise estimate (52) for the composite kernel K and the half-space Navier-Stokes theory that the later kernel bounds build on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the pointwise estimates (38) for the kernels G* and Q and the Stokes representation theory on which the whole estimate chain rests."},{"cited_title":"Maremonti, V","cited_arxiv_id":null,"evidence_quote":"Supplies the iterative scheme, the generalized Gronwall lemma, and the uniqueness strategy adapted here to the half-space."},{"cited_title":"Solonnikov,Estimates of the solutions of the nonstationary Navier-Stokes system, Zap","cited_arxiv_id":null,"evidence_quote":"Gives the existence and representation theory for the nonstationary Stokes and Navier-Stokes systems and the algebraic lemma used to close the iteration."},{"cited_title":"Crispo and P","cited_arxiv_id":null,"evidence_quote":"Provides the absolute-continuity property (P) used to obtain the t tending to 0 limit and the qualitative dependence of T(u0) on the data."},{"cited_title":"Foias, Une remarque sur l’unicit´ e des solutions des ´ equations de Navier-Stokes en dimensionn","cited_arxiv_id":null,"evidence_quote":"Supplies the duality argument at the core of the uniqueness proof in Theorem 2."}],"review_version":1}