{"id":"718379fc-75b7-4a06-95af-657220a527c5","arxiv_id":"2502.10024","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Continuation of 2D inhomogeneous Euler solutions follows from controlling ∂_{∇^⊥ρ}u, requiring only directional, not full, velocity gradient control.","lead":"A new proof shows that solutions of the 2D density-dependent incompressible Euler equations can be continued as long as certain derivatives of the velocity are controlled only along the direction tangent to density level lines. This weakens the classical whole-gradient continuation condition and recovers global well-posedness for constant density as a trivial case.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the proof is internally consistent; the no-vacuum assumption is a stated scope restriction rather than a hidden gap.","rationale":"I read the paper as a conditional continuation result: under (A1)-(A4), finite lifespan implies geometric norm blow-up. I walked through the two key chains. In the subcritical case, Lemma 3.1 converts the geometric condition (7) into a logarithmic bound on ||∇u||_{L∞}, and Proposition 3.3 plus Osgood gives uniform N(t); the pressure estimates in Lemma 3.2 are standard. In the critical case, the reduction to Proposition 4.1 is valid: condition (9) gives (20), so the Section 3 bounds apply; the B^0_{∞,1} estimates via Theorem 2.13 for X and η are correct; the difficult product ∂X u·u is handled by Lemma 2.5 and Proposition 2.6 without a loss of derivatives, and the Jensen step leading to U' ≤ C(1+U log(e+U)) is sound. I checked the algebraic identities for X and η and the paraproduct decompositions (44) and those leading to (51)-(55); they are consistent. The only place where the argument would break is if ρ is allowed to vanish; that is exactly the assumption (A2) the paper states. Hence the reader's ACCEPT verdict stands, with the same caveat.","tokens_in":29162,"tokens_out":32040,"duration_ms":293321,"concrete_test":"Run the Section 3 a priori estimates for the family of initial data ρ0^ε = ρ0 + ε, with ε>0, and track the implied constants in (32), (35), and (39) as ε→0. In particular, inspect inequality (30): if the best constant C(ε) in ||u||_{L∞} ≤ C(ε)||m||_{L∞} blows up like ε^{-1}, then the no-vacuum hypothesis is genuinely load-bearing and the criterion cannot be extended to densities with a vacuum set by a simple limiting argument. Conversely, if uniform-in-ε bounds are obtained, the theorem likely extends to ρ0≥0.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is a continuation/blow-up criterion proved under (A2), the no-vacuum condition 0<ρ_*≤ρ0≤ρ*. That hypothesis is used essentially: (i) to divide by ρ in the pressure equation (34) and in the elliptic estimate for ΔΠ; (ii) to convert bounds on m=ρu into bounds on u (step (30) uses ||u||_{L∞} ≲ ||m||_{L∞}); (iii) to propagate B^0_{∞,1} estimates for ∇ρ and η via (49)-(50), where the transport field u must be Lipschitz. If ρ0 touches zero, these steps stop being valid, so the stated criterion is not proved in the vacuum case. This is not an internal inconsistency: the theorem explicitly imposes (A2). The remaining structure—η=curl(ρu), the transport equations (23) and (26), the logarithmic Besov interpolation Lemmas 2.5 and 2.6, and the final Osgood-type bound on U—reads as consistent, and I did not find a gap in the estimates (45)-(57).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper establishes geometric blow-up and continuation criteria for the 2-D incompressible Euler equations with variable density, under the no-vacuum condition 0<ρ*≤ρ0≤ρ*. The main result (Theorem 1.3) states that a finite lifespan forces the L1-in-time norm of ∂_{∇⊥ρ}u to blow up in the subcritical Besov case, and the limsup of the B0_{∞,1} norm of ∂_{∇⊥ρ}u to blow up in the critical case. The proof is by contraposition: under suitable control of only the directional derivative along X=∇⊥ρ, the solution is continued past the given time. The novelty is to study the equation for X=∇⊥ρ together with a new unknown η=curl(ρu), which satisfies a transport equation whose right-hand side is exactly ∂_X u·u. The critical case relies on improved transport estimates in B0_{∞,1} and two new logarithmic interpolation inequalities.","tokens_in":29383,"tokens_out":21083,"duration_ms":211932,"significance":"If correct, this is a substantial advance: it replaces the known continuation condition ∫‖∇u‖_{L∞}<∞ by control of ∇u only along one geometric direction, namely the tangent direction to the density level sets. The proof is well structured: equations (10) and (11) make the geometric quantity appear naturally, the subcritical argument is clean and uses an Osgood-type bound, and the critical argument is nontrivial because B0_{∞,1} is not an algebra. The paper also provides useful new interpolation lemmas in logarithmic Besov spaces. The no-vacuum condition is explicitly stated and used in a transparent way, so the scope restriction is clear and not a hidden flaw. The dependence on Proposition 1.2 and Lemma 3.12 of the author's earlier work [5] is also explicitly identified.","major_comments":[],"minor_comments":[{"comment":"The passage from the estimate containing ‖u‖_{L∞} log(e+U′(τ)/‖u‖_{L∞}) to Eq. (56) is too terse. The monotonicity of β(z)=z log(e+c/z) alone does not directly remove the factor ‖u‖_{L∞}; one should add the elementary bound ‖u(t)‖_{L∞} log(e+U′(τ)/‖u(t)‖_{L∞}) ≤ C(1+log(e+U′(τ))), which follows from the uniform upper bound in (32). Since the inequality is elementary and the constant can be absorbed in the already implicit constants, this is a clarification rather than a conceptual gap.","section":"Section 4.3, around Eq. (56)"},{"comment":"The sentence 'Theorem 1.9 in fact implies Theorem 1.8' is logically inaccurate, since condition (9) is stronger than condition (8). What is meant is that the proof of Theorem 1.9 can be adapted to prove Theorem 1.8; please rephrase to avoid confusion.","section":"Section 4, first paragraph"},{"comment":"There is a typo in the heading 'Bouns for the non-linear term ∂_X u·u'; it should read 'Bounds for the non-linear term ∂_X u·u'. There are also minor spelling issues such as 'simplfy' in the same section.","section":"Section 4.3"},{"comment":"The symbol 'Tℓif e' appears garbled in several places (e.g., the introduction and Section 1.2); this is likely a font/encoding issue and should be fixed to 'T_life' or the intended notation.","section":"Introduction and notation"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is sound and the main results are significant. The issues I found are local and do not affect the validity of the central claim. I recommend publication after the clarifications above are made; the reliance on [5] is acceptable given that the relevant statements are precisely quoted and the current paper verifies their hypotheses."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper delivers what it promises: a blow-up/continuation criterion for 2-D non-homogeneous incompressible Euler that replaces control of the full velocity gradient with control of the single directional derivative along X = ∇⊥ρ. That is a real advance, not a repackaging. The new unknown η = curl(ρu) kills the pressure and turns the vorticity equation into a clean transport equation, and the two logarithmic Besov interpolation lemmas (Lemma 2.5 and Proposition 2.6) are genuinely new and do real work in the critical case. The proof is long and technical, but the core identities (10), (11), (23), (26) check out, and the bootstrap/Osgood argument is coherent. I found no fitted parameters, no circularity, and no hidden assumption that the theorem doesn't state. The recovery of global well-posedness for constant density as the degenerate case X ≡ 0 is a nice sanity check, not a gimmick.\n\nThe soft spots are real but proportionate. The no-vacuum hypothesis (A2) is used essentially—dividing by ρ, the Lax–Milgram pressure estimate, converting L∞ bounds on m to bounds on u, and propagating the B^0_{∞,1} estimates all need the density bounded away from zero. The paper states this clearly, so it is a scope restriction, not a flaw, but readers hoping for a vacuum-inclusive result will be disappointed. Second, the argument leans on Proposition 1.2 and Lemma 3.12 from the author's earlier work [5] as black boxes, and a few steps are compressed as 'standard.' That is acceptable in this subfield, but a referee should ask for those dependencies to be spelled out more explicitly. Third, in the critical case the cleanest statement, Theorem 1.9, needs a pointwise-in-time bound on the B^0_{∞,1} norm of ∂X u, not just an integrable one; the paper explains why this is the price of avoiding derivative loss, and I think the explanation is honest.\n\nWho is this for? People working on continuation criteria for inhomogeneous fluids, Besov-space well-posedness, and geometric/striated regularity. It is a solid, citable contribution. I would send it to a serious referee with confidence that the referee's time will be well spent; I would not desk-reject it. The main request should be to expand the compressed black-box estimates and to state plainly how far the no-vacuum assumption can be relaxed—probably not at all with the present method.","headline":"A genuinely new geometric continuation criterion for 2-D inhomogeneous Euler, built on a clever new unknown and two useful interpolation lemmas; the proof is sound and the only real caveat is the explicit no-vacuum scope.","tokens_in":29816,"tokens_out":1431,"would_cite":true,"duration_ms":16206,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q31","35B60","76B70"],"pacs":[],"model":"deepseek-v4-flash","headline":"A single directional derivative along density level lines controls blow-up for 2-D density-dependent Euler flows.","keywords":["incompressible Euler equations","variable density","blow-up criterion","continuation criterion","directional derivative","Besov spaces","logarithmic Besov spaces","two-dimensional fluid dynamics"],"falsifier":"Any concrete smooth solution with no vacuum and finite lifespan whose directional norms remain bounded would falsify the criteria: in the subcritical case, $\\int_0^{T^*}\\|\\partial_{X(t)}u(t)\\|_{L^\\infty}\\,dt<\\infty$ despite $T^*<\\infty$; in the critical case, $\\limsup_{t\\to T^*}\\|\\partial_{X(t)}u(t)\\|_{B^0_{\\infty,1}}<\\infty$. Numerically, this amounts to tracking $\\partial_X u$ along a family of solutions approaching a suspected singularity and checking whether the directional norm diverges at the blow-up time.","tokens_in":28996,"feed_emoji":"🌊","tokens_out":8026,"duration_ms":74352,"temperature":0.7,"pith_summary":"This paper establishes geometric blow-up and continuation criteria for the two-dimensional incompressible Euler equations with variable density, a system for which global well-posedness is not known. The central claim is that, instead of controlling the full gradient of the velocity $u$, it suffices to control the derivative of $u$ along the single direction $X=\\nabla^\\perp\\rho$, the tangent to the level lines of the density $\\rho$. In subcritical Besov regularity, integrability of $\\|\\partial_X u\\|_{L^\\infty}$ in time guarantees continuation; at critical regularity, a pointwise bound on $\\|\\partial_X u\\|_{B^0_{\\infty,1}}$ does the same without losing derivatives. The proof introduces a new unknown, the momentum vorticity $\\eta=\\mathrm{curl}(\\rho u)$, whose transport equation is free of the pressure and driven exactly by $\\partial_X u\\cdot u$. Since $X\\equiv 0$ for constant density, the classical global well-posedness of homogeneous 2-D Euler is recovered as a special case.","feed_headline":"Blow-up is decided by one directional derivative of velocity","feed_subtitle":"For 2-D variable-density Euler flow, controlling only the derivative along density level lines decides blow-up.","key_machinery":"The two ingredients that carry the argument are the transported vector field $X=\\nabla^\\perp\\rho$, which satisfies $\\partial_t X+u\\cdot\\nabla X=\\partial_X u$, and the momentum vorticity $\\eta=\\mathrm{curl}(\\rho u)$, which satisfies $\\partial_t\\eta+u\\cdot\\nabla\\eta=\\partial_X u\\cdot u$. Taking the curl of the momentum equation removes the pressure, so regularity of $\\eta$ can be propagated from the geometric quantity $\\partial_X u$. The reconstruction of $\\nabla u$ uses the Helmholtz decomposition of the momentum $m=\\rho u$, with singular integrals controlled in $B^0_{\\infty,1}$; logarithmic Besov spaces supply interpolation inequalities that avoid any loss of derivatives in the critical case, while improved transport estimates give bounds on $X$ and $\\eta$ growing only linearly with the Lipschitz norm of $u$.","core_discovery":"On the paper's own terms, the discovery is Theorem 1.3: if a smooth solution of the density-dependent Euler system on $\\mathbb{R}^2$ has finite lifespan $T^*<\\infty$, then, with $X(t)=\\nabla^\\perp\\rho(t)$, the subcritical case must satisfy $\\int_0^{T^*}\\|\\partial_{X(t)}u(t)\\|_{L^\\infty}\\,dt=\\infty$, and the critical case must satisfy $\\limsup_{t\\to T^*}\\|\\partial_{X(t)}u(t)\\|_{B^0_{\\infty,1}}=\\infty$. This is proved by contraposition as continuation criteria (Theorems 1.7 to 1.9): control of only the directional derivative along $X$ up to time $T$ lets the solution be extended beyond $T$ with the same Besov regularity. Because $X$ is transported by the flow and vanishes when $\\rho$ is constant, the result also confirms that singularities cannot form inside regions of constant density.","pith_inferences":["Beyond the paper: the same pressure-free momentum-vorticity trick may transfer to other 2-D inhomogeneous fluid models, such as damped or slightly viscous variants, where a similar transport equation for $\\mathrm{curl}(\\rho u)$ holds; this is not established here.","The paper leaves open whether a single scalar quantity, a true Beale-Kato-Majda analogue, could replace $\\partial_X u$ entirely; a first test would be whether controlled vorticity $\\omega$ and controlled density gradients can still produce a singularity.","The no-vacuum assumption is intrinsic to the method: if solutions with vacuum regions are allowed, the stated criteria are not proven, and a different geometric mechanism may be needed."],"forward_implications":["A subcritical solution continues past time $T$ as soon as $\\int_0^T\\|\\partial_{X(t)}u(t)\\|_{L^\\infty}\\,dt<\\infty$.","A critical solution continues past $T$ under the pointwise bound $\\sup_{t\\in[0,T)}\\|\\partial_{X(t)}u(t)\\|_{B^0_{\\infty,1}}<\\infty$, with no loss of derivatives.","At critical regularity, the weaker integral condition involving both $\\partial_X u$ and $\\partial_X |u|^2$ in $B^0_{\\infty,1}$ also guarantees continuation.","For constant density, $X\\equiv 0$, so the criteria never trigger and the homogeneous 2-D Euler equations are globally well-posed.","Any finite-time singularity must drive the directional derivative along the density's level lines to blow up, so regions of constant density cannot host singularities."],"supporting_citations":[{"why":"Provides the $L^p$ Besov well-posedness framework and pressure estimates used in the subcritical continuation proof.","marker":"[15]"},{"why":"Extends well-posedness to endpoint Besov spaces, the setting for the critical-case theorems.","marker":"[16]"},{"why":"Supplies the general continuation criterion in $L^1_T(L^\\infty)$ that the critical proof reduces to, plus pressure estimates reused here.","marker":"[5]"},{"why":"Gives the improved transport estimates with linear growth of $B^0_{p,r}$ norms used to propagate $X$ and $\\eta$.","marker":"[26]"},{"why":"Provides the companion improved transport estimate in borderline Besov spaces invoked for $B^0_{\\infty,1}$ bounds.","marker":"[21]"},{"why":"Supplies Littlewood-Paley calculus, paraproduct estimates, the logarithmic interpolation inequality, and transport theorems.","marker":"[2]"},{"why":"Introduces logarithmic Besov spaces, the finer interpolation framework used to avoid loss of derivatives in the critical case.","marker":"[17]"},{"why":"Introduces the paraproduct decomposition used to handle products such as $\\partial_X u\\cdot u$ in $B^0_{\\infty,1}$.","marker":"[4]"}],"fun_headline_variants":["One directional derivative decides Euler blow-up","Blow-up in 2-D Euler pinned to a single gradient","A lone derivative controls variable-density blow-up","Sharper blow-up test: just one velocity derivative","Density-level derivative gates Euler blow-up"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the density is bounded away from zero everywhere ($0<\\rho_*\\le\\rho_0\\le\\rho^*$); the proof divides by $\\rho$, solves the pressure equation by Lax-Milgram, recovers $u$ from $\\rho u$, and propagates Besov bounds, and all of these steps fail if vacuum regions appear.","fun_headline_variants_meta":{"raw":{"variants":["One directional derivative decides Euler blow-up","Blow-up in 2-D Euler pinned to a single gradient","A lone derivative controls variable-density blow-up","Sharper blow-up test: just one velocity derivative","Density-level derivative gates Euler blow-up"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000194,"raw_usage":{"total_tokens":1340,"prompt_tokens":918,"completion_tokens":422,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":534,"completion_tokens_details":{"reasoning_tokens":350}},"tokens_in":534,"tokens_out":422,"duration_ms":5153,"temperature":1.0,"reasoning_tokens":350,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T19:41:39.563173+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Any concrete smooth solution with no vacuum and finite lifespan whose directional norms remain bounded would falsify the criteria: in the subcritical case, $\\int_0^{T^*}\\|\\partial_{X(t)}u(t)\\|_{L^\\infty}\\,dt<\\infty$ despite $T^*<\\infty$; in the critical case, $\\limsup_{t\\to T^*}\\|\\partial_{X(t)}u(t)\\|_{B^0_{\\infty,1}}<\\infty$. Numerically, this amounts to tracking $\\partial_X u$ along a family of solutions approaching a suspected singularity and checking whether the directional norm diverges at the blow-up time.","supporting_citations":[{"cited_title":"Danchin: On the well-posedness of the incompressible density-depen dent Euler equations in the Lp framework","cited_arxiv_id":null,"evidence_quote":"Provides the $L^p$ Besov well-posedness framework and pressure estimates used in the subcritical continuation proof."},{"cited_title":"Danchin, F","cited_arxiv_id":null,"evidence_quote":"Extends well-posedness to endpoint Besov spaces, the setting for the critical-case theorems."},{"cited_title":"Bravin, F","cited_arxiv_id":null,"evidence_quote":"Supplies the general continuation criterion in $L^1_T(L^\\infty)$ that the critical proof reduces to, plus pressure estimates reused here."},{"cited_title":"Vishik: Hydrodynamics in Besov spaces","cited_arxiv_id":null,"evidence_quote":"Gives the improved transport estimates with linear growth of $B^0_{p,r}$ norms used to propagate $X$ and $\\eta$."},{"cited_title":"Hmidi, S","cited_arxiv_id":null,"evidence_quote":"Provides the companion improved transport estimate in borderline Besov spaces invoked for $B^0_{\\infty,1}$ bounds."},{"cited_title":"Fourier analysis and nonlinear partial differential equa- tions","cited_arxiv_id":null,"evidence_quote":"Supplies Littlewood-Paley calculus, paraproduct estimates, the logarithmic interpolation inequality, and transport theorems."},{"cited_title":"Mathematical analysis of models of non-homogeneous fluids and of hyperbolic operators with low-regularity coefficients","cited_arxiv_id":null,"evidence_quote":"Introduces logarithmic Besov spaces, the finer interpolation framework used to avoid loss of derivatives in the critical case."},{"cited_title":"Bony: Calcul symbolique et propagation des singularités pour les é quations aux dérivées par- tielles non linéaires","cited_arxiv_id":null,"evidence_quote":"Introduces the paraproduct decomposition used to handle products such as $\\partial_X u\\cdot u$ in $B^0_{\\infty,1}$."}],"review_version":1}