{"id":"f1eb820f-7af9-4376-896f-fc0e40b33781","arxiv_id":"2412.15625","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Free boundary incompressible magnetohydrodynamics is shown to be well-posed at the sharp low-regularity scale s > d/2 + 1, with first uniqueness and smooth-solution construction results in a new state space.","lead":"This paper proves a sharp existence, uniqueness, and continuous-dependence theory for magnetohydrodynamics with a free surface, that is, for a conducting liquid with a magnetic field whose boundary moves freely, in any dimension and on general domains. The result closes a completely open problem at low regularity and introduces a functional setting that may transfer to other free boundary models.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Critical cubic boundary estimate (4.11) is deferred to [18] and not verified for MHD; uniqueness and continuous dependence rest on an unproved transfer.","rationale":"The reader identified the Taylor sign condition as the weakest assumption and also flagged the omitted proof of the cubic boundary estimate. I agree with the latter as the most load-bearing concern because it is internal to the proof: the Taylor sign condition is a standard, physically motivated hypothesis, whereas the unverified transfer of [18, Eq. (4.11)] to the MHD setting is a specific technical step on which uniqueness and continuous dependence hinge. The manuscript itself acknowledges the omission, which strengthens the concern. The rest of the architecture—energy coercivity, linearized estimates, and the construction scheme—is plausible but cannot be fully assessed without the deferred calculations. The paper contains several other 'omit the proof' instances (e.g., Lemma 5.4, parts of Lemma 6.13 and Corollary 6.14), but (4.11) is the clearest single point where a failure would invalidate a main theorem. Thus the reader's CONDITIONAL verdict is appropriate and unchanged by this stress test; the concern reinforces the need for a complete, self-contained verification before the claim of sharp Hadamard well-posedness can be accepted.","tokens_in":77419,"tokens_out":3213,"duration_ms":27389,"concrete_test":"Independently re-derive (4.11) following [18, §4.4] but with the MHD pressure satisfying (1.5). Specifically, expand P−P_h on the intersection A using the pressure Poisson equation with the magnetic source term tr(∇B)^2 − tr(∇v)^2, and re-run the proof of [18, Eq. (4.11)] to see whether every step transfers. If any step uses only the Euler pressure structure, or if the magnetic contributions to P−P_h produce a term in the cubic boundary integral not controlled by the parameters A^{1/2} and A from (1.9)–(1.10), then Thm 1.6 and Thm 1.8 are unsupported. A minimal check is to compute the leading magnetic contribution to P−P_h near Γ∩Γ_h in the C^{1,1/2} regime and verify it is absorbed by the existing control parameters.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The stability theorem (Thm 1.8) and uniqueness theorem (Thm 1.6) both rely on estimate (4.11), which controls the cubic boundary integral ∫_A a^{-1}(P-P_h)(v-v_h)·∇(P-P_h) dS. The manuscript states on p. 23 that '(4.11) is far from trivial' and 'has exactly the same structure as the delicate cubic term in our previous work [18, Equation (4.11)]', leaving verification to the reader; Section 1.5.3 also says the proof is 'omitted from this manuscript'. This is load-bearing because Theorem 1.9's uniqueness and continuous dependence are obtained via Theorem 1.8, as stated in §1.4.2 and §7. The transfer to MHD is not automatic: the pressure now solves (1.5) with source tr(∇B)^2 − tr(∇v)^2, so P−P_h contains magnetically generated contributions absent in the Euler case. The claim that substituting p−p_h with P−P_h works requires a genuine check that the boundary-layer analysis in [18, §4.4] survives this extra nonlinearity. As written, the manuscript is not self-contained at a critical juncture of the proof.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript develops a well-posedness theory for the free boundary incompressible MHD equations (1.1)-(1.4) on bounded connected domains in arbitrary dimension. Its central claims are Theorem 1.6 (enhanced uniqueness under only W^{1,∞}-type control), Theorem 1.8 (stability of an L^2-level distance functional), Theorem 1.9 (Hadamard well-posedness in the state space H^s of Definition 1.2 for s > d/2 + 1, with continuous dependence and persistence of the Taylor sign condition), Theorem 1.11 (continuation criterion), and a new construction of smooth solutions (Section 6). The key innovations are an Elsässer-variable reformulation of MHD as two coupled free-boundary Euler-type systems, a state space incorporating the wave-type regularity condition ∇_B v, ∇_B B ∈ H^{s-1/2}, a normal-form-corrected energy functional, and an Eulerian time-discretization scheme with carefully designed regularizations. For B ≡ 0 the results recover the sharp Euler theorems of [18].","tokens_in":77470,"tokens_out":13731,"duration_ms":110142,"significance":"If correct, Theorem 1.9 resolves the low-regularity well-posedness of free boundary incompressible MHD on general domains, a genuinely open problem, and it does so without fitted parameters: the s > d/2 + 1 threshold matches the known sharp Euler threshold, and setting B ≡ 0 recovers known results as a consistency check. The linearized energy (Prop. 2.1), the explicit good-variable construction, and the 'regularizing effect' inequality (Lemma 5.12) are concrete, checkable contributions, and the existence scheme in Section 6 is an original route that avoids Lagrangian coordinates and simple-connectivity assumptions. The main reservation is verification: several load-bearing estimates, most importantly the cubic boundary estimate (4.11) behind uniqueness and continuous dependence, are explicitly left to the reader or deferred to the authors' companion paper [18], and the manuscript itself acknowledges this in §1.5.3 and §4.1.","major_comments":[{"comment":"The estimate (4.11) is load-bearing but is not proved. The manuscript states on p. 23 that \"(4.11) is far from trivial\" and that it \"has exactly the same structure as the delicate cubic term in our previous work [18, Equation (4.11)]\", and it \"leave[s] the verification of (4.11) to the reader\"; §1.5.3 adds that the proof is \"omitted from this manuscript\" and requires \"a subtle boundary layer analysis on the intersection of two domain states, which in general has only Lipschitz regularity.\" Theorem 4.1 is the engine behind Theorem 1.6 (uniqueness), Theorem 1.8 (stability), and the continuous-dependence conclusion of Theorem 1.9 (see §1.4.2 and §7), so the omitted estimate is load-bearing for the central claims. The transfer from [18] is not automatic: the MHD pressure solves (1.5) with source tr(∇B)^2 − tr(∇v)^2, so P − P_h contains magnetic contributions that have no Euler counterpart, and the boundary-layer analysis in [18, §4.4] would have to be re-verified against these terms; a check that only substitutes p − p_h by P − P_h is insufficient as written. Please include a proof of (4.11), or at least a detailed verification that the new quadratic magnetic source does not alter the structure of the estimate.","section":"§4.1, Eq. (4.11); §1.5.3"},{"comment":"Several lemmas on which the main proofs explicitly rely are deferred to [18] or \"left to the reader\". Lemma 5.4 supplies the bounds (4.4) used inside the proof of Theorem 4.1, and its proof reads \"entirely similar to [18, Lemmas 7.5 and 7.9], so we leave the details to the reader\"; Proposition 5.10 is invoked as \"a consequence of Proposition 7.14 in [18]\" with the remark that the proof \"applies almost verbatim\"; the surface-regularity control in §5.3.4 is declared \"virtually identical\" to [18, §7.4] with the proof omitted. A similar pattern recurs in Section 6 (e.g., Lemma 6.7 \"We omit the details\" and Corollary 6.14 \"left to the reader\"), where those computations feed the energy-monotonicity bound (6.21). Since the pressure now depends on B through the magnetic sources, and since the new state-space condition (iv) modifies the a priori regularity available in these estimates, the transfer of each item from the Euler case should be documented rather than asserted, even in a paper of this length.","section":"§5.3.1, §5.3.4, §5.4.1; §6.3"}],"minor_comments":[{"comment":"The regions A, A_h and the intersection hypersurface Γ̃ are central to the proof of Theorem 4.1, but their geometry (in particular the Lipschitz character of Γ̃ and the behavior of the weight b near Γ ∩ Γ_h) is described only in words; a figure or a short coordinate description would substantially aid the reader.","section":"§4.1, Eqs. (4.2)-(4.3)"},{"comment":"The notation O_{H^{k-3}}(τ^2) (used, for instance, in \"B̃_τ · n_τ = O_{H^{k-3}}(τ^2)\") and O_{C^3}(ε^2) is never defined; the subscripted big-O convention should be stated at first use.","section":"§6.1, §6.3"},{"comment":"The claim that the control parameter A_{1/2} in (1.10) is \"straightforward\" to control by ‖(v,B,Γ)‖_{H^s} for any s > d/2 + 1 is what places H^s solutions in the uniqueness class of Theorem 1.6, but no pointer to the relevant estimate is given; a reference to Corollary 5.6 together with the boundary Sobolev embedding would improve verifiability.","section":"§1.4.1"},{"comment":"There are several typographical errors: \"magnetohyrodynamics\" appears in the abstract and in the header, and \"satisifes\" appears in §6.3; these should be corrected.","section":"Abstract, p. 1; §6.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is ambitious and, if the deferred estimates hold, would be a landmark contribution to the free-boundary MHD literature. My recommendation of major revision is driven by the gap at (4.11) and by the heavy reliance on [18] for cornerstones of the proof; both are repairable within the current scope. Two further points for the editor: (i) the manuscript's own text in §1.5.3 and on p. 23 explicitly concedes that the proof of a key estimate is omitted, and such an admission should weigh materially in the decision process; (ii) the reliance on [18] for the functional framework (Section 3 and much of Appendix A are presented as recalled from that paper) makes the incremental novelty of the present work—while real—somewhat harder for a general reader to isolate. The historical and literature claims appear accurate, and the B ≡ 0 recovery provides a meaningful consistency check."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the first serious well-posedness theory for free boundary incompressible MHD at the sharp Sobolev scale s > d/2 + 1, in arbitrary dimension and for non-simply-connected domains. If the estimates hold, it settles a central open problem. The proof is not self-contained at one critical juncture, and that is where the weight sits.\n\nWhat is genuinely new: the state space includes the wave-type condition ∇_B v, ∇_B B ∈ H^{s−1/2}, which is the right way to capture the MHD coupling; the uniqueness theorem is the first at any regularity for general data; and the time-discrete construction of smooth solutions is a substantial technical innovation, not a repackaging. The paper is also unusually candid about its own limits: Remark 1.1, the explicit \"omitted from this manuscript\" note in Section 1.5.3, and the stated conjectures. No fitted parameters, and the B = 0 limit recovers the known Euler results, which is a good sanity check.\n\nNow the soft spot, which is exactly the one the stress-test flags. Estimate (4.11), controlling the cubic boundary integral, is load-bearing for the stability theorem and for uniqueness. It is explicitly deferred to [18] on page 23, with the paper calling it \"far from trivial.\" The transfer to MHD is not a formality: the pressure now solves (1.5) with a magnetic source tr(∇B)^2 − tr(∇v)^2, so P − P_h contains contributions absent in the Euler case. Saying the analysis \"carries over\" is not a proof. A referee needs to verify that the boundary-layer analysis in [18, Section 4.4] survives this extra nonlinearity. This is a real gap in self-containedness, not a stylistic concern.\n\nSecond, the later sections (6.6–6.8, 7, and Appendix A) are not fully accessible in the posted version I read. The construction of rough solutions and continuous dependence in H^s rests on those parts, so the paper is not fully checkable end-to-end from the arXiv text. That is a practical obstacle, not evidence of error.\n\nNone of this is fatal. The architecture is coherent: linearized energy, coercive energy with good variables and a normal-form correction, difference estimates, and a discrete iteration. The claims are precise. If (4.11) transfers, this is a major result; if it fails, the failure will teach us something real about the MHD problem.\n\nSend it to a serious referee. The referee should be asked specifically to verify the transfer of (4.11) and whether the new state space is genuinely propagated, not just assumed. This paper deserves referee time.","headline":"A genuinely new well-posedness framework for free boundary MHD at the sharp scale; the one load-bearing estimate is deferred to a previous paper, so referee scrutiny should focus there.","tokens_in":78207,"tokens_out":3557,"would_cite":true,"duration_ms":35554,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","76B03","76B15","76W05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Free boundary MHD is well-posed at the sharp Sobolev threshold s > d/2 + 1, in arbitrary dimensions and on general domains, with unique solutions depending continuously on the data.","keywords":["free boundary MHD","Hadamard well-posedness","Taylor sign condition","Elsässer variables","low-regularity Sobolev spaces","wave-type regularity","free boundary Euler equations","normal form correction"],"falsifier":"Compute (analytically or numerically) the difference functional D between two nearby solutions in the collar with identical initial data but with the magnetic field difference switched on; if the cubic boundary term (4.11), namely $\\int_A a^{-1}(P-P_h)(v-v_h)\\cdot\\nabla(P-P_h)\\,dS$, cannot be bounded by $C_A (A^{1/2}+A^{1/2}_h)D(W,W_h)$ uniformly at the stated low regularity, then uniqueness or continuous dependence fails. Concretely, find a sequence of states at $s = d/2+1+\\varepsilon$ whose Taylor coefficient is bounded below by $c_0$ but whose boundary develops a Lipschitz corner inside the collar; the proof's boundary-layer analysis must still control the pressure difference on the intersection $\\tilde{\\Gamma}$.","tokens_in":76996,"feed_emoji":"🌊","tokens_out":5251,"duration_ms":44030,"temperature":0.7,"pith_summary":"This paper claims a complete Hadamard well-posedness theory for the free boundary incompressible MHD equations in Sobolev spaces of order s > d/2 + 1, in any dimension and on general, not necessarily simply connected, domains. The central assertion is that a state (v, B, Γ) with velocity and magnetic field in H^s, a boundary in H^s, the Taylor sign condition, and the wave-type regularity $\\nabla_B v, \\nabla_B B \\in H^{s-1/2}$ evolves uniquely for a time depending only on the H^s size and the Taylor lower bound, with continuous dependence on the data. If correct, this settles the low-regularity well-posedness question that had been open for the free boundary MHD system. Taking B = 0 recovers the sharp free boundary Euler results.","feed_headline":"Free boundary MHD is well-posed at sharp low regularity","feed_subtitle":"A new state space with wave-type regularity yields existence, uniqueness, and continuous dependence for s > d/2 + 1.","key_machinery":"The load-bearing object is the new H^s state space whose norm includes $\\|\\nabla_B v\\|_{H^{s-1/2}} + \\|\\nabla_B B\\|_{H^{s-1/2}}$; this captures the wave operator $D^2_t - \\nabla_B^2$ hidden in the MHD system. The arguments are carried by Elsässer variables $W^\\pm = v \\pm B$; Alinhac-style good variables $G^\\pm = D^\\pm_t a - \\nabla_n \\Delta^{-1} D^\\pm_t \\Delta P$; a coercive energy functional $E_k \\approx M_{s-1/2} \\|(v,B,\\Gamma)\\|^2_{H^k}$ with rotational and irrotational parts; the regularizing effect $\\nabla_B^2 a \\in H^{k-2}(\\Gamma)$ that follows from the magnetic tangency condition; and a stability functional measuring the $L^{2}$ distance of $W^\\pm$ plus a weighted $L^{2}$ difference of pressures on the intersection of domains. Existence of regular solutions is built by a time-discretized Euler-plus-transport iteration with three regularization steps: parabolic surface and irrotational regularization, mild mollification, and elliptic regularization in the B direction.","core_discovery":"The paper's central discovery is that the right state space for the free boundary MHD equations includes, beyond the natural H^s regularity of v, B and Γ, the half-derivative wave-type condition $\\nabla_B v, \\nabla_B B \\in H^{s-1/2}$, and that this state space is dynamically propagated. Writing $W^\\pm = v \\pm B$ turns the system into two coupled free boundary Euler-like transport equations $D^\\pm_t W^\\mp = -\\nabla P$ with both $D^\\pm_t$ tangent to the free surface; the Taylor coefficient $a = -\\nabla P \\cdot n$ remains the good boundary variable, and the boundary evolution acquires the form $D^2_t a - \\nabla_B^2 a + a N a = f$. On this basis the paper proves uniqueness at essentially $W^{1,\\infty}$ control, stability via a nonlinear distance functional, local well-posedness at $s > d/2 + 1$, and a low-regularity continuation criterion.","pith_inferences":["Because the state-space condition is formulated in Eulerian coordinates and avoids Lagrangian flow-map regularity, the same triple-step regularization scheme may transfer to other free boundary models with wave-type coupling, such as plasma-vacuum interfaces or MHD with surface tension.","The omitted verification of the cubic boundary estimate (4.11) means the uniqueness and continuous-dependence conclusions currently rest on the transfer of a delicate estimate from the Euler case; a self-contained proof in the MHD setting would remove the main proof-level gap a reader cannot check in this manuscript.","One can test the robustness of the Taylor sign assumption by examining whether the constructed solutions break down when a0 touches zero; the paper cites ill-posedness in that regime, so the threshold c0 > 0 is expected to be sharp."],"forward_implications":["Any smooth solution whose Taylor coefficient stays bounded away from zero and whose low-regularity norm stays bounded can be continued past time T; this is the paper's Theorem 1.11.","Taking B = 0 recovers the sharp well-posedness result for free boundary Euler equations, so the MHD coupling does not degrade the regularity threshold.","The wave-type condition $\\nabla_B v, \\nabla_B B \\in H^{s-1/2}$ propagates from the initial data as part of the solution, rather than being an extra constraint imposed at every time.","Uniqueness holds in a much weaker class than existence: any two solutions with finite control parameters A and $A_{1/2}$ coincide.","The results hold in arbitrary dimensions and on not-necessarily-simply-connected domains, unlike previous existence proofs."],"supporting_citations":[{"why":"Supplies the Eulerian framework, the balanced elliptic estimates, and the cubic boundary estimate (4.11) whose verification is deferred to the reader.","marker":"[18]"},{"why":"Cited for ill-posedness of the free boundary MHD equations when the Taylor sign condition fails.","marker":"[16]"},{"why":"Classical ill-posedness result for free boundary Euler that motivates the Taylor sign condition.","marker":"[9]"},{"why":"The only prior general-domain result, giving local existence without uniqueness or continuous dependence in a high-regularity simply connected setting.","marker":"[23]"},{"why":"Time-discretization structure (regularization plus Euler iteration) that the construction of regular solutions adapts.","marker":"[19]"},{"why":"First a priori estimates for the free boundary MHD problem, under restrictive boundary conditions that later works relax.","marker":"[15]"}],"fun_headline_variants":["Free boundary MHD: sharp well-posedness achieved","Sharp low-regularity well-posedness for free boundary MHD","New Eulerian approach yields MHD free boundary theory","Complete Hadamard theory for MHD at s > d/2 + 1","MHD free boundary: well-posed at sharp Sobolev regularity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on the Taylor sign condition $a_0 = -\\nabla P_0 \\cdot n_{\\Gamma_0} > c_0 > 0$, meaning the total pressure is a non-degenerate defining function of the initial boundary; the paper cites ill-posedness when it fails, and every theorem invokes it, with persistence of the lower bound being part of the proof. A secondary proof-level premise is that the delicate cubic boundary estimate (4.11), whose verification is deferred to [18] and explicitly omitted here, carries over to the MHD difference functional.","fun_headline_variants_meta":{"raw":{"variants":["Free boundary MHD: sharp well-posedness achieved","Sharp low-regularity well-posedness for free boundary MHD","New Eulerian approach yields MHD free boundary theory","Complete Hadamard theory for MHD at s > d/2 + 1","MHD free boundary: well-posed at sharp Sobolev regularity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00031,"raw_usage":{"total_tokens":1817,"prompt_tokens":1046,"completion_tokens":771,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":662,"completion_tokens_details":{"reasoning_tokens":679}},"tokens_in":662,"tokens_out":771,"duration_ms":4312,"temperature":1.0,"reasoning_tokens":679,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:16:04.561799+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute (analytically or numerically) the difference functional D between two nearby solutions in the collar with identical initial data but with the magnetic field difference switched on; if the cubic boundary term (4.11), namely $\\int_A a^{-1}(P-P_h)(v-v_h)\\cdot\\nabla(P-P_h)\\,dS$, cannot be bounded by $C_A (A^{1/2}+A^{1/2}_h)D(W,W_h)$ uniformly at the stated low regularity, then uniqueness or continuous dependence fails. Concretely, find a sequence of states at $s = d/2+1+\\varepsilon$ whose Taylor coefficient is bounded below by $c_0$ but whose boundary develops a Lipschitz corner inside the collar; the proof's boundary-layer analysis must still control the pressure difference on the intersection $\\tilde{\\Gamma}$.","supporting_citations":[{"cited_title":"Sharp Hadamard local well-posedness, enhanced uniqueness and pointwise continuation criterion for the incompressible free boundary Euler equations","cited_arxiv_id":"2309.05625","evidence_quote":"Supplies the Eulerian framework, the balanced elliptic estimates, and the cubic boundary estimate (4.11) whose verification is deferred to the reader."},{"cited_title":"Ill-posedness of free bounda ry problem of the incompressible ideal MHD","cited_arxiv_id":null,"evidence_quote":"Cited for ill-posedness of the free boundary MHD equations when the Taylor sign condition fails."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Classical ill-posedness result for free boundary Euler that motivates the Taylor sign condition."},{"cited_title":"On the Free Boundary Problems for the Ideal Incompressible MHD Equations","cited_arxiv_id":"2311.06581","evidence_quote":"The only prior general-domain result, giving local existence without uniqueness or continuous dependence in a high-regularity simply connected setting."},{"cited_title":"The compressible Eule r equations in a physical vacuum: A comprehensive Eulerian approach","cited_arxiv_id":null,"evidence_quote":"Time-discretization structure (regularization plus Euler iteration) that the construction of regular solutions adapts."},{"cited_title":"A priori estimates for free bo undary problem of incompressible inviscid magnetohydrody - namic flows","cited_arxiv_id":null,"evidence_quote":"First a priori estimates for the free boundary MHD problem, under restrictive boundary conditions that later works relax."}],"review_version":1}