{"id":"1c8c3886-5681-405f-9409-a75000d0b20c","arxiv_id":"2411.17153","paper_version":4,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A low-regularity Hadamard-style local well-posedness theorem is proved for the full compressible Euler equations with a physical vacuum boundary in all space dimensions.","lead":"This paper proves that the non-isentropic compressible Euler equations in a physical vacuum are locally well-posed for rough initial data whose gas-vacuum interface may have unbounded curvature. It provides existence, uniqueness, and continuous dependence on initial data in weighted Sobolev spaces, closing a gap left by existing isentropic and Lagrangian analyses.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The uniform-lifespan bootstrap in §6.2 requires 'B0 chosen so that B0 ≫ A*0 and B0 ≫ M0', but B0 is fixed by the initial data; the proof as written does not cover data with small B0 but large H2κ norm.","rationale":"The paper's Eulerian reformulation, weighted spaces, and the architecture borrowed from Ifrim-Tataru appear coherent, and I did not find an internal contradiction in the main energy estimates themselves. The most load-bearing issue is the unclosed bootstrap in §6.2: the text explicitly treats B0 as a quantity that can be chosen, although B0 is determined by the initial datum. Since this bootstrap is what converts high-regularity solutions into uniform-in-h bounds and then into rough solutions, the existence and continuous-dependence claims of Theorem 1.3 depend on closing it. The proposed high-frequency example tests whether the closure can be achieved without imposing an impossible condition on B0. Because the flaw is in the proof as written rather than in a demonstrated falsehood of the theorem, a conditional acceptance is appropriate: the authors should either give a data-dependent choice of T0 that closes the bootstrap, or restrict the statement to data satisfying the extra condition. The reader's weakest-assumption analysis focused on the physical-vacuum scaling and entropy bounds, which is a different concern; hence partial rather than full agreement.","tokens_in":74,"tokens_out":29166,"duration_ms":512965,"concrete_test":"Take d ≥ 2, κ > κ0 + 1/2, set v0 = 0, σ0 = 1, and q0 = r + δ φ(x) sin(h·x), where r is a nondegenerate defining function, φ is supported away from Γ0, and δ = h^{-2κ}. Verify that ∥(q0,0,1)∥_{H2κ_{q0}} = O(1), while A*0 and B0 tend to 0 as h → ∞. Then attempt to close the §6.2 bootstrap with T0 = 1/(C(B0,c0)). If the closure used to improve (6.7) requires the displayed condition B0 ≫ M0, the argument stalls; if it only requires a sufficiently small T0, display the choice of T0 in terms of A*0, B0, M0, and c0 and verify all bootstrap improvements. This isolates whether the proof establishes uniform lifespans for all H2κ data.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In the proof of the rough-solution existence theorem, after deriving the bound (6.13), the text states: 'In particular, if B0 is chosen so that B0 ≫ A*0 and B0 ≫ M0, the above estimates and the Sobolev embeddings will improve the bootstrap assumptions for Bh.' Here B0 is the control parameter (1.17) of the given initial datum, not a free parameter. Definition 1.2 does not constrain A*0, B0, and M0 = ∥(q0,v0,σ0)∥_{H2κ} to satisfy B0 ≫ M0; indeed, one can have A*0 and B0 tending to zero while M0 remains order one by adding to a smooth defining function a high-frequency, small-amplitude oscillatory perturbation with frequency h and amplitude h^{-2κ}. The bootstrap assumptions (6.7) are used to obtain (6.8)-(6.13) and then to improve Bh, Ah*, and c^h. If the improvement genuinely requires B0 ≫ A*0 and B0 ≫ M0, the existence theorem fails for exactly the low-regularity data with large high-energy but small low-order controls. If that condition is replaceable by an explicit smallness condition on T0 depending on A*0, B0, M0, and c0, the replacement is not written, leaving the central bootstrapping step incomplete in the current text.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript develops an Eulerian-coordinate theory for the non-isentropic compressible Euler equations with a physical vacuum, using the variables (q, v, σ) introduced in (1.5)-(1.6). Its central result, Theorem 1.3, asserts Hadamard-style local well-posedness in weighted Sobolev spaces H^{2κ} for any κ > κ0 + 1/2 = 1 + d/2 + 1/(2β), allowing gas-vacuum interfaces with unbounded curvature. The paper also proves a uniqueness theorem in a Lipschitz-type regularity class (Theorem 1.1 and the quantitative Theorem 3.1), sharp a priori energy estimates (Theorem 1.5), and a continuation criterion (Theorem 1.4). The method follows the Eulerian scheme of Ifrim-Tataru: linearized energy estimates, weighted interpolation and coercivity arguments, Euler-polygonal construction of high-regularity solutions via a two-scale regularization, and finally frequency-envelope/interpolation arguments for rough solutions.","tokens_in":63568,"tokens_out":12631,"duration_ms":124371,"significance":"If the main theorem is correct, this is a substantial advance: it extends the isentropic Eulerian physical-vacuum theory of Ifrim-Tataru [20] to variable entropy, while providing continuous dependence, a priori estimates, and continuation criteria in a low-regularity setting. The reformulation that reduces the fully nonlinear entropy coupling to bilinear terms is a genuine step, and the a priori estimates are derived from the equations rather than fitted to data. The uniqueness theorem covers almost all classical solutions. The main reservation is a bootstrap gap in §6.2 that affects the proof of existence in Theorem 1.3 for a non-negligible part of the stated data class; this is a load-bearing issue rather than a presentation concern.","major_comments":[{"comment":"The bootstrap closure for the lifespan of the regularized solutions is not valid for the full data class stated in Theorem 1.3. The text says 'if B0 is chosen so that B0 ≫ A∗0 and B0 ≫ M0', but B0 is not a free parameter: it is the control parameter (1.17) of the given initial datum, and Definition 1.2 imposes no relation between B0, A∗0, and M0 = ∥(q0,v0,σ0)∥_{H^{2κ}}. Data with small B0 and order-one M0 are easy to construct: fix a smooth φ supported away from Γ0 and add a_N φ(Nx) to v0 or σ0 with a_N = N^{-2κ}; then M0 remains of order one while ∥∇(·)∥_{L∞} + ∥·∥_{C^{1/2+ε}} tends to zero because κ > κ0 + 1/2 > 1/2. For such data the bootstrap assumptions (6.7), the estimate (6.13), and the claimed improvement do not close, so the uniform lifespan T independent of h is not established and Theorem 1.3 is not proved as stated. The proof needs either a bootstrap whose constants are functions of M0 (with T0 also allowed to depend on M0 and c0), or an explicit smallness/relation condition on the data; as written, the existence theorem is incomplete.","section":"§6.2 (after Eq. (6.13))"}],"minor_comments":[{"comment":"The displayed physical energy Ephy lacks the volume element dx, and the following sentence 'one can first infer from (1.9) that ∫ ... dx' is grammatically incomplete; the intended conclusion 'is conserved' should be written out.","section":"§1.3, Eq. (1.9)"},{"comment":"The notation κ := σ1 + σ2 in (3.6) collides with the Sobolev regularity index κ used throughout the paper; renaming the sum (for example Σ) would remove a source of confusion.","section":"§3.1, Eq. (3.6)"},{"comment":"The last displayed condition in (5.25) mixes ε and ϵ in the term '(ϵ2t)2χ2ε', and the left-hand side '(εt)χε · (1 − χε)' would be clearer with a consistent product notation.","section":"§5.5.2, Eq. (5.25)"},{"comment":"The statement introduces δ as an arbitrarily small constant in the envelope estimate and later uses δ and δ′ for a boundary-layer separation parameter; these uses should be distinguished notationally.","section":"§6.1, Proposition 6.2"},{"comment":"Several load-bearing technical tools (Propositions 4.1–4.4, 5.4, and 6.2) are quoted from [20] without proof. This is acceptable as a citation practice, but the manuscript should explicitly flag at each first use that these are imported results and state the exact hypotheses under which they are being invoked.","section":"§4–§6, imported results"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically dense and the non-isentropic extension is potentially significant. My main concern is the §6.2 bootstrap issue, which is not merely cosmetic: it affects the uniform-lifespan argument in the proof of the central Theorem 1.3. I would be willing to accept a revised version that supplies a correct bootstrap closure (for example, with constants depending on the full H^{2κ} norm) or that precisely restricts the data class."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is my take on Liu-Luo. The paper proves low-regularity Hadamard LWP for non-isentropic compressible Euler in a physical vacuum, in Eulerian coordinates, allowing the interface to have unbounded curvature. That is a genuine open problem, and the main theorems are exactly the right targets. The reformulation with q and σ is natural: it turns the fully nonlinear momentum equation into a bilinear one, and the entropy-weighted Sobolev spaces are well motivated by conservation laws. The uniqueness theorem in a Lipschitz class and the a priori energy estimates are substantial, and the details in Sections 3–5 are coherent. The reliance on Ifrim–Tataru for the interpolation, regularization, and frequency-envelope lemmas is acceptable if those are indeed published and correct; the citations are proper, though a referee will need to trust or re-check them.\n\nThe soft spot is in Section 6.2. The bootstrap for the regularized solutions uses the sentence \"if B0 is chosen so that B0 ≫ A*0 and B0 ≫ M0...\". But B0 is not a free parameter; it is a fixed quantity of the initial data. There is no argument covering data with small B0 but large H2κ norm, for instance a smooth defining function plus a high-frequency bump of small amplitude. Such data satisfy the hypothesis of Theorem 1.3 but not the condition used in the bootstrap. As written, the proof therefore does not establish existence for exactly the low-regularity data it claims to cover. This may be repairable by choosing bootstrap constants depending on M0 and letting T0 depend on M0 as well, but that replacement is not written. This is a load-bearing gap, not a cosmetic one. I would rate soundness lower than the reader's 6—more like a 4 or 5 until the gap is fixed.\n\nOverall: the reformulation and the high-regularity existence are valuable, and the rough-solution existence has a real hole. This paper deserves a serious referee, but the authors should be asked to close the bootstrap gap before the main theorem can be accepted.","headline":"Serious extension of Ifrim-Tataru to non-isentropic Euler, but the rough-solution bootstrap in §6.2 has a gap that excludes data with small B0 and large H2κ norm.","tokens_in":64138,"tokens_out":5379,"would_cite":false,"duration_ms":51337,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","35Q31","76N10","35L60"],"pacs":[],"model":"deepseek-v4-flash","headline":"The full compressible Euler equations with a physical vacuum are locally well-posed in low-regularity weighted Sobolev spaces, in every space dimension, even when the gas-vacuum interface has unbounded curvature.","keywords":["compressible Euler equations","physical vacuum","free boundary problems","local well-posedness","weighted Sobolev spaces","non-isentropic gas dynamics","continuation criteria","Eulerian coordinates"],"falsifier":"Construct a non-isentropic initial state in d=2 with \\$\\beta$=1, an interface with unbounded curvature, and \\kappa just above 1 + \\frac12 + \\frac12, then run the one-step polygonal iteration at decreasing time steps: if the discrete energy increments ever exceed (1+C\\varepsilon) times the initial energy, the a priori estimate underlying the existence proof fails. Alternatively, exhibit two distinct $H^{{2\\kappa}}$ limiting solutions from the same initial data, which would refute the uniqueness claim.","tokens_in":1613,"feed_emoji":"🌬️","tokens_out":6575,"duration_ms":108625,"temperature":0.7,"pith_summary":"This paper claims that the full non-isentropic compressible Euler equations with a physical vacuum are locally well-posed in low-regularity weighted Sobolev spaces, in any space dimension, for initial data so rough that the gas-vacuum interface may have unbounded curvature. In other words, existence, uniqueness, and continuous dependence on the data all hold for such rough states, a question that was previously open for high-dimensional non-isentropic flows. The paper also proves sharp a priori energy estimates and a continuation criterion that guarantees the flow can be extended as long as the boundary stays non-degenerate and certain control parameters remain bounded. The proof works in Eulerian coordinates, avoiding the regularity loss of the flow map that would come with Lagrangian coordinates.","feed_headline":"Rough gas-vacuum interfaces admit unique local Euler flows","feed_subtitle":"Existence, uniqueness, and continuous dependence hold even when the interface has unbounded curvature.","key_machinery":"The argument is carried by a reformulation and by adapted 'good unknowns'. The change of variables q = \\frac{1+\\$\\beta$}{\\$\\beta$} $p^{{\\beta/(1+\\beta)}}$ and \\$\\sigma$ = $e^{{S/(1+\\beta)}}$ turns the pressure gradient into a bilinear term, giving the system D_t q + \\$\\beta$ q \\nabla\\cdot v = 0, D_t v + \\$\\sigma$ \\nabla q = 0, D_t \\$\\sigma$ = 0. Local well-posedness is then proved inside weighted Sobolev spaces built from powers of q, using self-adjoint degenerate-elliptic operators such as L_1 u = \\$\\beta$ q \\$\\Delta$ u + \\nabla q\\cdot\\nabla u and L_2 w = \\$\\beta$ \\nabla(q\\nabla\\cdot w) + (\\nabla w)^*\\nabla q to construct coercive energy functionals. The good unknowns s_{2k} = $D_t^{{2k}}$q - \\nabla q\\cdot $D_t^{{2k-1}}$v and w_{2k} = $D_t^{{2k}}$v satisfy linearized systems with admissible error terms, which keeps the energy propagation estimates free of derivative loss. Existence for rough data is obtained by Euler's polygonal iteration with carefully bi-scale regularized data, and the passage to fractional regularity uses frequency envelopes and interpolation.","core_discovery":"The central claim is Theorem 1.3: the full compressible Euler system written in the new variables (q, v, $\\sigma$) is locally well-posed in the state space $H^{{2\\kappa}}$ for every real \\kappa > \\kappa_0 + \\frac12 = 1 + \\frac d2 + \\frac{1}{2\\$\\beta$}, where d is the space dimension and \\$\\beta$ is the polytropic exponent. In these spaces, q, v, and $\\sigma$ are controlled by weighted Sobolev norms with weights $q^{{\\kappa+(\\alpha-1)/2}}$, $q^{{\\kappa+\\alpha/2}}$, and $q^{{\\kappa+\\alpha/2}}$, where q is a renormalized effective pressure and \\$\\alpha$ = \\$beta^{{-1}}$. The Sobolev embeddings then imply that the free boundary is at least $C^{{1.5+}}$ and may have unbounded curvature, so the result covers interfaces outside the classical smooth-boundary theory. The paper establishes solutions in C([0,T]; $H^{{2\\kappa}}$), uniqueness in a larger Lipschitz-type solution class, weak Lipschitz dependence on the data in an $L^{2}$-type distance, continuous dependence in the $H^{{2\\kappa}}$ topology, sharp a priori energy estimates, and a continuation criterion.","pith_inferences":["Because the entropy variable \\sigma simply transports along particle paths, a similar Eulerian treatment may extend to other systems with transport coefficients, such as Euler-Poisson equations or multi-species gas models, once a renormalized conserved quantity with a uniformly bounded inverse is identified; this is beyond what the paper proves.","The threshold \\kappa > 1 + \\frac d2 + \\frac{1}{2\\beta} suggests that the entropy adds roughly half a derivative of difficulty: the most singular new terms involve \\nabla\\sigma weighted by powers of q, so pushing the regularity lower would likely require a different treatment of entropy waves.","The bi-scale regularization construction described in the paper could be read as a blueprint for numerical schemes: low-scale regularized data combined with high-scale corrections reproduces the energy without accumulating derivative loss, making the polygonal iteration a plausible computational strategy."],"forward_implications":["Free gas-vacuum interfaces with unbounded curvature evolve deterministically for a short time, since unique local solutions now exist for such rough initial data.","The a priori energy estimate gives a Gronwall bound with no loss of derivatives, so the lifespan of a solution depends only on the size of the initial state and the control parameters.","Solutions can be continued as long as the boundary remains non-degenerate, \\inf_{\\Gamma_t}|\\nabla q| \\geq c_0 > 0, and the control parameters A_* and B remain bounded as required; in particular, splash-type singularities are excluded while these conditions hold.","Continuous dependence on initial data holds in the H^{2\\kappa} topology, so nearby initial states produce nearby solutions in a concrete, checkable sense.","The same Eulerian framework works in all space dimensions d \\geq 1 and for every polytropic exponent \\beta > 0."],"supporting_citations":[{"why":"Supplies the Eulerian framework, the weighted Sobolev spaces, the good unknowns, and the bi-scale regularization method that the paper adapts to the non-isentropic case.","marker":"[20]"},{"why":"Shows ill-posedness for sound-speed exponents other than the physical-vacuum scaling, justifying the restriction to the physical-vacuum regime.","marker":"[24]"},{"why":"Established local well-posedness for the isentropic physical vacuum in 1D, the first such result that this work extends to non-isentropic flows.","marker":"[22]"},{"why":"Established local well-posedness for the isentropic physical vacuum in 3D, providing the main comparison baseline for the Lagrangian framework.","marker":"[25]"},{"why":"Established uniqueness for classical solutions in 3D with bounded curvature, which the present uniqueness theorem improves by allowing unbounded curvature and lower regularity.","marker":"[39]"},{"why":"Introduces frequency envelopes, which the paper uses to characterize fractional-order weighted Sobolev spaces and to obtain continuous dependence.","marker":"[49]"}],"fun_headline_variants":["Euler well-posedness extends to rough gas-vacuum interfaces","Unbounded curvature allowed: Euler local well-posedness","Low-regularity Euler solutions on curved vacuum boundaries","Physical vacuum dynamics: existence, uniqueness for rough interfaces","Local Euler flows exist beyond smooth-boundary theory"],"cache_read_input_tokens":66176,"weakest_assumption_plain":"The load-bearing assumption is the physical-vacuum boundary scaling: the sound speed squared must vanish like the distance to the gas-vacuum interface, with the renormalized entropy bounded above and below, so that q has nonzero gradient at the boundary; if the sound speed decayed with a different power of distance, the weighted Sobolev spaces would not be adapted and the problem is expected to be ill-posed.","fun_headline_variants_meta":{"raw":{"variants":["Euler well-posedness extends to rough gas-vacuum interfaces","Unbounded curvature allowed: Euler local well-posedness","Low-regularity Euler solutions on curved vacuum boundaries","Physical vacuum dynamics: existence, uniqueness for rough interfaces","Local Euler flows exist beyond smooth-boundary theory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000411,"raw_usage":{"total_tokens":2104,"prompt_tokens":897,"completion_tokens":1207,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":1126}},"tokens_in":513,"tokens_out":1207,"duration_ms":10713,"temperature":1.0,"reasoning_tokens":1126,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T12:27:06.351026+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a non-isentropic initial state in d=2 with \\$\\beta$=1, an interface with unbounded curvature, and \\kappa just above 1 + \\frac12 + \\frac12, then run the one-step polygonal iteration at decreasing time steps: if the discrete energy increments ever exceed (1+C\\varepsilon) times the initial energy, the a priori estimate underlying the existence proof fails. Alternatively, exhibit two distinct $H^{{2\\kappa}}$ limiting solutions from the same initial data, which would refute the uniqueness claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Eulerian framework, the weighted Sobolev spaces, the good unknowns, and the bi-scale regularization method that the paper adapts to the non-isentropic case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows ill-posedness for sound-speed exponents other than the physical-vacuum scaling, justifying the restriction to the physical-vacuum regime."},{"cited_title":"Pure Appl","cited_arxiv_id":null,"evidence_quote":"Established local well-posedness for the isentropic physical vacuum in 1D, the first such result that this work extends to non-isentropic flows."},{"cited_title":"Pure Appl","cited_arxiv_id":null,"evidence_quote":"Established local well-posedness for the isentropic physical vacuum in 3D, providing the main comparison baseline for the Lagrangian framework."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Established uniqueness for classical solutions in 3D with bounded curvature, which the present uniqueness theorem improves by allowing unbounded curvature and lower regularity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces frequency envelopes, which the paper uses to characterize fractional-order weighted Sobolev spaces and to obtain continuous dependence."}],"review_version":1}