{"id":"c2eaf331-08c8-4089-a51f-85a32da9a6d7","arxiv_id":"2411.17929","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Forced Oberbeck-Boussinesq equations with a Newtonian gravity term have infinitely many weak solutions with the same zero initial data.","lead":"This paper claims that forced Oberbeck-Boussinesq equations with gravity admit infinitely many weak solutions from zero initial data on a very short time interval. It builds on the known unstable profile of the Navier-Stokes equations, but the proof as written has several gaps that need repair.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.1 is proved for an operator with the opposite drift sign to the one used in the contraction, so the temperature semigroup estimate that closes the fixed-point argument is not established as written.","rationale":"The reader's weakest assumption correctly identifies Lemma 3.1 as the load-bearing point, and my reading of the full text confirms the sign inconsistency: the operator L is defined with +Ū·∇, but the equation that the proof analyzes and the linearized temperature equation contain −Ū·∇. This is not a cosmetic typo, because the Duhamel formula (3.3) uses the stated L; if one does not flip the sign, the proved semigroup estimate does not apply to the object in the contraction. The scaling error in the proof of Lemma 3.1 further undermines the derivation as printed. The τ0 condition is a second, independent flaw: the contraction constants are small only for τ0 → −∞, contrary to the text's 'τ0 sufficiently small'. None of these issues suggests the theorem is false; the underlying strategy follows the established unstable-profile construction of Albritton–Brué–Colombo, and the gravitational term is handled by the Hardy inequality after projecting away the gradient contribution. The flaws are localized and likely repairable, so the verdict should remain CONDITIONAL rather than moving to REJECT or ACCEPT. My agreement with the reader is full: the same concern was identified, and the proposed concrete test would settle whether the sign error is merely typographical or reflects a deeper obstruction.","tokens_in":77,"tokens_out":22520,"duration_ms":437829,"concrete_test":"Rewrite §3.2–§3.4 with the single convention L = ∆ + 1/2(1+ξ·∇) − Ū·∇, the operator appearing in the linearized temperature equation, and re-derive the Duhamel formula (3.3) and estimates (3.12)–(3.15) using the corrected scaling ‖D^k ū‖_{L∞} ∼ t^{-(k+1)/2}. If the estimates hold unchanged, the gap is a sign typo and the contraction can be repaired by taking τ0 negative with large magnitude; if any displayed inequality changes or fails, the existence of Θ_p is not established by the current argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The contraction argument for the temperature perturbation relies entirely on the semigroup estimate for L. In §3.1 the linearized and perturbation equations for Θ contain the operator ∂τ − ∆ − 1/2(1+ξ·∇) + Ū·∇, so the generator is L_correct = ∆ + 1/2(1+ξ·∇) − Ū·∇. However, the paper defines L = ∆ + 1/2(1+ξ·∇) + Ū·∇ and states Lemma 3.1 for this operator, while the proof of Lemma 3.1 actually treats the equation ∂τΘ − ∆Θ − 1/2(1+ξ·∇)Θ = −Ū·∇Θ, i.e. the operator with minus drift. Thus the lemma, as proved, does not control the semigroup used in (3.3); the Duhamel formula for Θ_p is not justified by the stated estimate. The proof of Lemma 3.1 also contains an incorrect scaling: for ū(x,t)=t^{-1/2}Ū(x/√t) one has ‖D^k ū‖_{L∞} ∼ t^{-(k+1)/2}, not t^{(k+1)/2}; the displayed estimate therefore does not follow from the preceding Leibniz-rule step. Separately, the contraction section claims Φ maps B to itself when τ0 is sufficiently small, but the printed bounds contain factors like e^{(2a−β)τ0} and e^{(γ−β)τ0}, which are small only when τ0 is sufficiently negative, not when it is close to zero. These are repairable in principle, but as written the key estimate is internally inconsistent and the fixed-point step is not certified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs non-unique weak solutions to the three-dimensional Oberbeck-Boussinesq system with Newtonian gravitational field and zero initial data, with external force and heat source in L^1_t L^2_x. The construction follows the unstable-profile strategy of Albritton, Brué, and Colombo for forced Navier-Stokes: in self-similar variables the solution is split into an unstable profile, an unstable linear mode, and a perturbation, the force and heat source are solved from the profile, and the perturbation is obtained by a contraction mapping. The main novelty is to absorb the drift term \\bar U·∇ into the heat semigroup for the temperature equation, leading to Lemma 3.1, which is then used to close the fixed-point argument for the temperature perturbation.","tokens_in":10104,"tokens_out":13856,"duration_ms":122076,"significance":"If Theorem 1.1 can be established, it would be a meaningful extension of the known non-uniqueness phenomenon for forced Navier-Stokes equations to the coupled Oberbeck-Boussinesq system with a Newtonian gravitational potential. The strategy of incorporating the drift term into the semigroup is natural, and the paper correctly identifies the temperature equation as the part where the standard heat-semigroup control is insufficient. The manuscript is not circular: it imports the unstable profile and semigroup estimates from [3] and constructs the forces explicitly. However, the key semigroup estimate for the operator actually used in the contraction is not established as written, so the current proof does not certify Theorem 1.1.","major_comments":[{"comment":"The sign of the drift term in the operator L is inconsistent. The equation solved by Θ_p is ∂_τ Θ_p − ∆Θ_p − 1/2(1+ξ·∇)Θ_p + \\bar U·∇Θ_p + (remaining nonlinear terms) = 0, so the relevant generator is ∆ + 1/2(1+ξ·∇) − \\bar U·∇. The paper defines L = ∆ + 1/2(1+ξ·∇) + \\bar U·∇, but the proof of Lemma 3.1 verifies the estimate for the solution of ∂_τ Θ − ∆Θ − 1/2(1+ξ·∇)Θ = −\\bar U·∇Θ, which is the operator with the opposite drift sign. Thus the semigroup estimate proved in §3.2 does not apply to the operator named L in (3.3), and the Duhamel representation of Θ_p is not justified by the stated lemma.","section":"§3.1, Eq. (3.3), and Lemma 3.1"},{"comment":"The scaling of the derivatives of \\bar u is incorrect. For \\bar u(x,t) = t^{-1/2}\\bar U(x/√t), one has ‖D^k \\bar u‖_{L∞} ≍ t^{-(k+1)/2}, not t^{(k+1)/2} as written. Consequently the displayed Leibniz-rule estimate involving the factor t^{-1} does not follow from the preceding line, and the estimate leading to (3.5) is not established. This is a separate defect from the sign issue above, and it concerns a load-bearing step in the proof of Lemma 3.1.","section":"§3.2, proof of Lemma 3.1"},{"comment":"The statement that Φ maps B to itself and is a contraction when τ0 is sufficiently small is not supported. Under the parameter conditions (3.16), the exponents 2a−β, a+b−γ, −β+min{a+β,2β,γ}, and −γ+min{a+γ,β+b,β+γ} are positive, so the displayed factors e^{(2a−β)τ0}, e^{(a+b−γ)τ0}, and the corresponding M-factors are small only when τ0 is sufficiently negative with large magnitude, not when τ0 is close to zero. Since t0 = e^{τ0}, this means t0 should be chosen very small, but the argument must be rewritten to require a sufficiently negative τ0 and to verify that all estimates are uniform on (−∞, τ0].","section":"§3.4, contraction mapping"}],"minor_comments":[{"comment":"The displayed definition of Φ1 appears with an integral from τ to ∞, whereas the Duhamel formula (3.2) and the estimates (3.6)–(3.10) use integration from −∞ to τ. As printed, the map does not match the intended fixed-point equation; this should be corrected.","section":"§3.4, definition of Φ1"},{"comment":"The displayed estimate for the term Up·∇Θp is missing the factor ‖Up‖_X ‖Θp‖_Y; as written the line contains only the exponential factor and so cannot imply the stated contraction bound.","section":"§3.4, Eq. (3.15)"},{"comment":"Theorem 1.1 uses g for the heat source, while system (1.1) and the rest of the paper use h; this should be harmonized.","section":"Theorem 1.1 and §3"},{"comment":"The definition of φ_k(t) omits the volume element dx, and several integration-by-parts steps, including the term (−1)^k ∫ ∆^k θ · ∂_t θ dx, are not explained; these should be written out or justified.","section":"§3.2, proof of Lemma 3.1"},{"comment":"The paper should specify that the compactly supported profile Θ(ξ) in \\bar Θ(ξ,τ)=e^{bτ}Θ(ξ) vanishes near ξ=0, or otherwise justify that the term \\bar Θ ∇(1/|ξ|) belongs to L^2_x pointwise in time, since the Hardy-inequality step in §3.3 is only applied to Θ_p.","section":"§3.3, choice of \\bar Θ"},{"comment":"There are several typographical errors and misspellings, for example 'sysem' in the abstract, 'Boussiness' in the introduction, 'Nest' on page 4, and 'τ′ is sufficiently small' in the last line of §3.4 instead of τ0; these should be corrected.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The result is potentially interesting and fits a PDE journal, and I do not see a circularity problem: the unstable profile and semigroup bounds are taken from [3], and the forces are solved explicitly. The main concern is the proof of Lemma 3.1, which currently contains both a sign inconsistency with the operator used in (3.3) and an incorrect scaling estimate; the contraction argument is therefore not certified as written. These issues appear repairable within the manuscript's scope, so I recommend major revision rather than rejection. The author should rewrite §3.2 and §3.4 carefully, state the correct sign of L, prove Lemma 3.1 for that operator, and replace the 'τ0 sufficiently small' statements with the required large-magnitude negative τ0 condition."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you my read on Liu's arXiv:2411.17929. The headline: it genuinely extends the non-uniqueness of forced Navier-Stokes to the Oberbeck-Boussinesq system with a Newtonian gravitational force. That is a real step in the current program, and it's the first time I've seen the unstable-profile construction applied to a coupled temperature-convection system.\n\nWhat is good: the overall strategy is clear. The author splits the solution into a self-similar profile, the unstable linear mode from [3], and a perturbation, then solves for the forces explicitly. The forces are not fitted to the conclusion. The technical novelty is the semigroup estimate for the temperature equation, which is the right idea for taming the convection term. The citation practice is honest: the paper builds directly on Albritton-Brué-Colombo and says so.\n\nThe soft spots are real. Lemma 3.1, the load-bearing estimate, is not established as written. The operator L is defined with a plus drift, but the equation actually solved in the proof has a minus drift. The proof treats the minus-drift operator, so the lemma does not cover what is used in (3.3). That looks like a sign typo, but it has to be fixed. More seriously, the scaling statement is wrong: for \\bar{u}(x,t)=t^{-1/2}U(x/\\sqrt{t}), D^k\\bar{u} decays like t^{-(k+1)/2}, not grows. That error undermines the derivation around (3.5). And the contraction step claims τ0 should be small, while the printed bounds only get small when τ0 is very negative (i.e., t0<1). The integration limits in the definition of Φ also appear flipped in Section 3.4. None of these look irreparable, but they are not cosmetic.\n\nBottom line: the theorem is plausible and the strategy is sound, but the current manuscript does not certify the proof. I would send it to a serious referee, because the question is important enough and the gap looks fixable. The author needs a careful rewrite of Section 3.2 and the parameter conditions before this goes anywhere.","headline":"Genuine extension of Navier-Stokes non-uniqueness to Boussinesq with gravitational forcing, but the key semigroup estimate has sign and scaling errors; worth a referee, not acceptable as-is.","tokens_in":10,"tokens_out":4326,"would_cite":false,"duration_ms":159012,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35L71","35B40","37K40"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper constructs force and heat-source terms for which the Oberbeck-Boussinesq system with Newtonian gravity has infinitely many weak solutions from zero data.","keywords":["Oberbeck-Boussinesq system","non-uniqueness","weak solutions","Navier-Stokes unstable profile","self-similar coordinates","contraction mapping","gravitational field","L1_t L2_x forcing"],"falsifier":"A direct check is to take a simple Gaussian profile $\\bar U$ and initial temperature $\\Theta_0$, write the linearized temperature equation in the paper's self-similar coordinates, and test numerically whether $\\|e^{\\tau L}\\Theta_0\\|_{H^k}$ grows at most like $\\max\\{\\tau^{-(k-m)/2},1\\}\\|\\Theta_0\\|_{H^m}$ and whether the apparent drift term has the sign the paper assigns it. If the estimate fails for one such profile, Lemma 3.1 and the contraction step fail with it.","tokens_in":9531,"feed_emoji":"🌡️","tokens_out":12812,"duration_ms":99948,"temperature":0.7,"pith_summary":"The paper proves a non-uniqueness theorem for the Oberbeck-Boussinesq system with a Newtonian gravitational field on $\\mathbb{R}^3$: there is a short time interval and external forces in $L^1_t L^2_x$ for which the system has infinitely many distributional solutions starting from rest. If correct, this shows that coupling the forced Navier-Stokes equations to a temperature equation and a gravitational term does not restore uniqueness in the weak class. The proof transplants an unstable mode of the linearized Navier-Stokes operator into self-similar coordinates, adds a non-steady temperature profile, and closes with a contraction mapping. A reader should care because the result makes weak solutions of thermal convection genuinely ambiguous, not merely formally underdetermined.","feed_headline":"Forced Boussinesq flow from rest has infinitely many weak solutions","feed_subtitle":"Same force and heat source can yield many temperature fields, so weak solutions do not settle convection.","key_machinery":"The load-bearing object is the linearized temperature semigroup $e^{\\tau L}$ with $L=\\Delta+\\tfrac12(1+\\xi\\cdot\\nabla)+\\bar U\\cdot\\nabla$ in self-similar variables. Lemma 3.1 claims it satisfies the heat-type smoothing estimate $\\|e^{\\tau L}\\Theta_0\\|_{H^k}\\lesssim \\max\\{\\tau^{-(k-m)/2},1\\}\\|\\Theta_0\\|_{H^m}$ for $k\\ge m\\ge0$, which is what lets the convection terms be absorbed in the contraction argument. The second ingredient is the unstable mode of the Navier-Stokes linearized operator $L_{ss}$: a smooth divergence-free profile with a positive eigenvalue and a parabolic semigroup bound, which seeds the velocity perturbation and forces the exponential weights in the solution spaces.","core_discovery":"Theorem 1.1 asserts that there exists $t_0>0$ and forces $f,h\\in L^1_t L^2_x((0,t_0)\\times\\mathbb{R}^3)$ such that (1.1) has infinitely many distributional solutions $(u,\\theta)\\in L^2_t L^2_x((0,t_0)\\times\\mathbb{R}^3)$ with initial data $(u_0,\\theta_0)=(0,0)$. The construction linearizes the self-similar form of the system around a compactly supported unstable profile $\\bar U$ of the Navier-Stokes linearized operator, chooses the linear velocity perturbation as the real part of the unstable mode $e^{\\lambda\\tau}\\rho$, sets a temperature profile $\\bar\\Theta(\\xi,\\tau)=e^{b\\tau}\\Theta(\\xi)$, and defines the force and heat source by substitution. The remaining perturbation is found by a contraction map, and multiplying the unstable mode by different constants yields infinitely many distinct solutions.","pith_inferences":["A natural next test is whether the same construction works for a bounded gravity vector such as $(0,0,1)$; the proof uses the Hardy inequality for $1/|\\xi|$, and a bounded $\\nabla G$ would need a different estimate for the temperature-to-velocity coupling.","The theorem does not address the unforced case $f=h=0$; here the forces are chosen to cancel the profile, so an unforced non-uniqueness, if possible, likely needs a new mechanism.","If the key temperature estimate is repaired, it could serve as a general tool for active-scalar systems with drift, making the whole strategy a template for transporting Navier-Stokes non-uniqueness into other coupled fluid models.","Because the solutions are built from an exponentially growing unstable mode, they probably violate the energy inequality; a testable conjecture is that energy-admissible weak solutions of this system remain unique even though the broader weak class is not."],"forward_implications":["If the theorem is correct, the forced Oberbeck-Boussinesq system on $\\mathbb{R}^3$ is non-unique in the weak class for zero initial data: the same force and heat source admit infinitely many distributional solutions.","The temperature field is not determined by the data in that class, so weak formulations of thermal convection need extra admissibility criteria to be predictive.","By the scaling property of the system, the construction gives examples on arbitrarily short time intervals, not just one fixed $t_0$.","The proof identifies a reusable mechanism: an unstable Navier-Stokes mode drives the linear velocity, and a heat-like semigroup estimate for the temperature equation with drift closes the contraction.","The result extends the known non-uniqueness of forced Navier-Stokes to a coupled system with gravity and thermal transport."],"supporting_citations":[{"why":"Supplies the unstable divergence-free profile, the positive eigenvalue, and the parabolic semigroup bound (Lemmas 2.1 and 2.3) that seed the whole construction.","marker":"[3]"},{"why":"Prior work on local-in-space behavior of Navier-Stokes whose self-similar unstable profile the paper treats as the starting point; the introduction presents the theorem as extending this line.","marker":"[18]"},{"why":"Companion study of ill-posedness in the energy space where the unstable profile was identified, cited together with [18] as the origin of the profile.","marker":"[19]"},{"why":"The well-posedness result for homogeneous initial data in the Oberbeck-Boussinesq system with Newtonian gravity; it is the baseline uniqueness statement that Theorem 1.1 contradicts and motivates the choice $G=1/|x|$.","marker":"[7]"}],"fun_headline_variants":["Even from rest, Boussinesq flow branches infinitely","Same heat and force, infinite weak solutions from rest","Zero initial data, infinitely many convection solutions","Gravity-driven Boussinesq: non-unique from rest","Infinite weak solutions for forced Boussinesq at rest"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on a single estimate: that the temperature equation keeps its heat-like smoothing property when the unstable background flow is added. The demonstration of that estimate in the paper is internally inconsistent — the drift term has conflicting signs, the claimed scaling of the background flow's derivatives does not match the coordinate change, and the parameter choices for the contraction step point the wrong way — so the theorem stands or falls on whether that estimate can be fixed.","fun_headline_variants_meta":{"raw":{"variants":["Even from rest, Boussinesq flow branches infinitely","Same heat and force, infinite weak solutions from rest","Zero initial data, infinitely many convection solutions","Gravity-driven Boussinesq: non-unique from rest","Infinite weak solutions for forced Boussinesq at rest"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000238,"raw_usage":{"total_tokens":1419,"prompt_tokens":759,"completion_tokens":660,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":375,"completion_tokens_details":{"reasoning_tokens":580}},"tokens_in":375,"tokens_out":660,"duration_ms":6803,"temperature":1.0,"reasoning_tokens":580,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:45:09.876464+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct check is to take a simple Gaussian profile $\\bar U$ and initial temperature $\\Theta_0$, write the linearized temperature equation in the paper's self-similar coordinates, and test numerically whether $\\|e^{\\tau L}\\Theta_0\\|_{H^k}$ grows at most like $\\max\\{\\tau^{-(k-m)/2},1\\}\\|\\Theta_0\\|_{H^m}$ and whether the apparent drift term has the sign the paper assigns it. If the estimate fails for one such profile, Lemma 3.1 and the contraction step fail with it.","supporting_citations":[{"cited_title":"Albritton, E","cited_arxiv_id":null,"evidence_quote":"Supplies the unstable divergence-free profile, the positive eigenvalue, and the parabolic semigroup bound (Lemmas 2.1 and 2.3) that seed the whole construction."},{"cited_title":"Jia and V","cited_arxiv_id":null,"evidence_quote":"Prior work on local-in-space behavior of Navier-Stokes whose self-similar unstable profile the paper treats as the starting point; the introduction presents the theorem as extending this line."},{"cited_title":"Jia and V","cited_arxiv_id":null,"evidence_quote":"Companion study of ill-posedness in the energy space where the unstable profile was identified, cited together with [18] as the origin of the profile."},{"cited_title":"Large self-similar solutions to Oberbeck-Boussinesq system with Newtonian gravitational field","cited_arxiv_id":"2311.01093","evidence_quote":"The well-posedness result for homogeneous initial data in the Oberbeck-Boussinesq system with Newtonian gravity; it is the baseline uniqueness statement that Theorem 1.1 contradicts and motivates the choice $G=1/|x|$."}],"review_version":1}