{"id":"9b870498-eb69-4cc9-85f0-0c2fe5644d4b","arxiv_id":"2507.16436","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":3,"one_line_summary":"A proof is proposed that small L1∩L2 initial data yields global classical solutions and optimal time-decay rates when the viscosity exponent α is within δ of 1.","lead":"The paper claims global solutions and optimal decay rates for the 3D compressible Navier-Stokes equations with density-dependent viscosity, assuming small L1∩L2 initial data and viscosity exponent α close to 1. Its advertised advance over prior work is that high-order Sobolev norms of the initial data need not be small.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bootstrap criterion (3.1) is imposed at t=0, but the hypothesis (1.7) does not imply the required L∞-smallness; estimates (3.42) and (3.51) silently drop high-derivative factors, so the continuity argument in §4.1 cannot be initialized.","rationale":"The paper is a serious attempt to relax the smallness assumptions for density-dependent compressible Navier-Stokes equations, and the Green's function plus bootstrap structure is clearly presented. However, the single most load-bearing point is the initialization of the bootstrap: Proposition 3.1 assumes (3.1), and the proof of Theorem 1.1 in Section 4.1 defines T* through (3.1) on [0,T]. At t=0, (3.1) is exactly an L∞-and-gradient smallness condition on the initial data. The theorem's hypothesis (1.7) only gives L1∩L2 smallness and explicitly allows arbitrarily large Sobolev norms of derivatives, so it does not imply the condition. The displayed estimates (3.42) and (3.51) are where the missing control appears: they use Sobolev interpolation to bring in ∥∇²V0∥ or ∥∇³V0∥ and then omit those factors, effectively assuming boundedness of high frequencies. This is not merely an aesthetic weakness; it is the step that would supply the initial upper bound needed for the bootstrap. A concrete admissible initial datum with a tall narrow density bump makes (3.1) false at t=0, so T*=0 and the continuity argument fails. The result may be recoverable by adding an L∞ smallness assumption on the initial data or by proving a separate short-time argument, but such a modification would weaken the advertised novelty that only L1∩L2 smallness is needed. Therefore the reader's REJECT verdict is appropriate, with the caveat that the identified gap is in the proof as written rather than a demonstrated counterexample to the theorem itself.","tokens_in":21073,"tokens_out":7761,"duration_ms":89222,"concrete_test":"Choose u0=0 and ϱ0(x)=Mφ(x/ε) with a smooth compactly supported bump φ and ∫φ=1. For M=ε^{-3/4} and ε small, ∥ϱ0∥_{L1} ≈ ε^{9/4} and ∥ϱ0∥_{L2} ≈ ε^{3/4}, so (1.7) holds; meanwhile ∥ϱ0∥_{L∞}=M=ε^{-3/4} and ∥∇ϱ0∥_{L∞} ≈ ε^{-7/4}, while the H^4 norm grows like ε^{-13/4}. This is exactly the large-derivative regime allowed by Remark 1.2. Evaluating (3.1) at t=0 gives 1/2 < ε^{-3/4} and η < ε^{-7/4}, so the bootstrap hypothesis is violated and the T* defined in §4.1 is zero. Repeating the first inequality of (3.42) for this datum also shows that the omitted factor ∥∇²V0∥_{L2}^{3/4} ≈ ε^{-15/16} grows, so the bound C C0^{1/4} is invalid. This explicit construction settles that the proof's continuity argument cannot start for admissible initial data.","verdict_should_be":"REJECT","load_bearing_attack":"Proposition 3.1 assumes the time-decay regularity criterion (3.1), and Section 4.1 defines T* = sup{T | (3.1) holds on [0,T]}. At t=0 this requires ∥(ϱ0,u0)∥_{L∞} ≤ 1/2 and ∥∇(ϱ0,u0)∥_{L∞} ≤ η. The theorem only assumes (1.7), ∥(ϱ0,u0)∥_{L1∩L2} ≤ C0, with H^4 data whose spatial derivatives may be arbitrarily large (Remark 1.2). L1∩L2 smallness does not control L∞ or ∇L∞ when high derivatives are large. In (3.42) the proof writes ∥V0∥_{L∞} ≤ C∥V0∥_{L2}^{1/4}∥∇²V0∥_{L2}^{3/4} and then replaces this by C C0^{1/4}, suppressing the potentially large factor ∥∇²V0∥_{L2}^{3/4}; (3.51) similarly drops ∥∇³V0∥_{L2}^{5/6}. Thus, for admissible data with a large narrow density bump, (3.1) is false at t=0, T* = 0, and the bootstrap never starts. This is a load-bearing gap in the proof of the central claim, not a matter of taste: unless (3.1) is restricted to t>0 and a separate short-time argument supplies it, none is given.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the three-dimensional isentropic compressible Navier-Stokes equations with density-dependent viscosities μ=ρ^α, λ=ρ^α near the constant state. Theorem 1.1 claims global existence of classical solutions and optimal decay rates for H^4 initial data under only smallness of the L1∩L2 norm of the perturbation and |α−1| small, with no smallness imposed on derivatives (Remark 1.2). The proof combines a Green's function decomposition (low-frequency heat-like part, regular high-frequency part, and singular part) with conditional energy estimates under a time-decay bootstrap assumption (3.1), an auxiliary L^{4/3} estimate, and a continuity argument in Section 4. The advertised novelty is that the Sobolev norms of derivatives of the initial data may be arbitrarily large.","tokens_in":21332,"tokens_out":13309,"duration_ms":122775,"significance":"If valid, the result would be a substantial improvement over the classical Matsumura-Nishida and Guo-Wang frameworks and over the recent Luo-Yang result for density-dependent viscosities, since it would remove derivative smallness. The paper contains a careful Green's function decomposition, detailed conditional energy estimates, and an interesting L^{4/3} auxiliary estimate. However, the bootstrap cannot be initialized under the stated hypotheses, and the linear estimates silently discard uncontrolled high-derivative factors. These issues concern the central claim, so the theorem is not established as written.","major_comments":[{"comment":"The bootstrap assumption (3.1) is required to hold on [0,T], hence in particular at t=0 it demands ∥(ϱ0,u0)∥_{L∞} ≤ 1/2 and ∥∇(ϱ0,u0)∥_{L∞} ≤ η. The hypotheses of Theorem 1.1 only assume (1.7), i.e. smallness in L1∩L2, and Remark 1.2 explicitly allows the Sobolev norms of derivatives to be arbitrarily large. Smallness in L1∩L2 does not imply either of the required L∞ bounds: for a bump ϱ0(x)=M φ(x/w) with fixed profile φ, one has ∥ϱ0∥_{L2}=M w^{3/2} and ∥ϱ0∥_{L1}=M w^3; with M=C0 w^{-3/2}, the L2 norm equals C0 and the L1 norm tends to 0 as w→0, while ∥ϱ0∥_{L∞}=C0 w^{-3/2} and ∥∇ϱ0∥_{L∞}∼C0 w^{-5/2} diverge. Consequently the quantity T* defined in §4.1 can be 0, and the continuity argument cannot be started. The proof needs either a separate short-time argument establishing (3.1) under (1.7) or a change of the main hypotheses; as written, the advertised result is not proven.","section":"§3.2 and §4.1, Eq. (3.1), (3.28), (1.7)"},{"comment":"In the linear estimate M1 in (3.42), the proof bounds ∥GHS(t)∗V0∥_{L∞} ≤ e^{-Ct}∥V0∥_{L∞} ≤ e^{-Ct}∥V0∥_{L2}^{1/4}∥∇²V0∥_{L2}^{3/4} and then replaces the right side by C C0^{1/4}(1+t)^{-3/2}. The factor ∥∇²V0∥_{L2}^{3/4} is dropped without justification. Similarly, in (3.51) the terms ∥∇V0∥_{L2} and ∥∇V0∥_{L∞} are interpolated as ∥V0∥_{L2}^{1/2}∥∇²V0∥_{L2}^{1/2} and ∥V0∥_{L2}^{1/6}∥∇³V0∥_{L2}^{5/6}, and the high-derivative factors are then replaced by constants C0^{1/2} and C0^{1/6}. Under the hypotheses of Theorem 1.1 these high-derivative norms are not controlled by C0; they may be arbitrarily large. These dropped factors are exactly what would make the linear terms small, so the closing of the bootstrap in Lemmas 3.6 and 3.7 is not justified.","section":"Eq. (3.42) and (3.51)"}],"minor_comments":[{"comment":"The statement says 'For 2 ≤ q ≤ ∞' but the estimate uses L^p; the index p is not defined and the relation between q and p is not given.","section":"Theorem 1.1, Eq. (1.11)"},{"comment":"In the K2 estimate, the displayed powers of C0 do not follow from (3.8): since (3.8) gives ∥V∥_{L2} ≤ C C0^{1/2}, the product ∥V∥_{L2}∥∇V∥_{L2} is O(C0^{11/10}) rather than O(C0^{8/5}). The exponents should be recomputed or clarified.","section":"§3.1, Lemma 3.5"},{"comment":"There are several typos: 'Naiver-Stokes' appears in §1 and in references [3,16]; 'Gargliardo-Nirenberg' appears in Lemma 3.5; 'The the general solution' appears in §2; and in (3.17) the sign of the (λ+μ)∇divu term differs from (1.6).","section":"§2 and §3"},{"comment":"The exponents α1 and α2 in (4.2) are asserted to be 'verified in Section 3.2', but no explicit verification is given; this should be supplied.","section":"§4.2"},{"comment":"The remark says that only optimal decay for ∥∇^k u∥_{L^2} is obtained, whereas Theorem 1.1 and §4.2 state decay for ∇^k(ϱ,u); the text should be reconciled.","section":"Remark 4.1"}],"recommendation":"reject","confidential_remarks":"The initialization gap concerns the central advertised novelty, namely the absence of smallness on derivatives. A revision that simply adds H^s smallness would not deliver the claimed theorem; a genuine repair requires a mechanism controlling high frequencies at t=0 that is not present in the manuscript. I recommend rejection, although the Green's function decomposition and the conditional energy estimates may be reusable in a corrected version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The central claim of this paper is not established. Proposition 3.1's hypothesis (3.1) is imposed for all t in [0,T], including t=0: it requires ∥(ϱ0,u0)∥L∞ ≤ 1/2 and ∥∇(ϱ0,u0)∥L∞ ≤ η. The theorem's only smallness condition is (1.7), smallness in L1∩L2, which does not control L∞ when higher derivatives are large. In (3.42) the proof interpolates ∥V0∥L∞ ≤ C∥V0∥_{L2}^{1/4}∥∇²V0∥_{L2}^{3/4} and then replaces that by C C0^{1/4}, dropping the potentially large high-derivative factor. The same happens in (3.51) with ∥∇³V0∥_{L2}^{5/6}. For admissible data with a tall narrow density bump, (3.1) is false at t=0, so T* in Section 4.1 is zero and the bootstrap never starts. This is a load-bearing gap, not a matter of taste.\n\nThat said, the paper is a serious attempt. The goal—removing Sobolev-norm smallness in favor of L1∩L2 smallness for density-dependent viscosities—is genuinely new relative to Luo–Yang [10] and Guo–Wang [4]. The conditional energy estimates in Lemmas 3.3 and 3.4 appear coherent under the regularity criterion, and the Green's function frequency decomposition is standard. The structure is clear, and the writing is readable despite a few typos.\n\nThe gap looks repairable: one could add a smallness assumption on the H^2 norm (or the L∞ norm) of the initial data, or supply a separate short-time argument that derives (3.1) from the H^4 regularity plus the energy bounds. But as written, the claimed theorem is not proved.\n\nThis paper is for researchers working on decay rates for the compressible Navier–Stokes equations. The claim is significant enough to warrant a serious referee, but the referee should be explicitly asked to check the initialization of the bootstrap and the interpolation inequalities in (3.42) and (3.51). I would send it to peer review, expecting major revision.","headline":"The main theorem is not proved as written: the bootstrap criterion (3.1) demands L∞ smallness at t=0, which the L1∩L2 assumption does not imply, and the linear estimates (3.42) and (3.51) silently drop high-derivative factors.","tokens_in":21953,"tokens_out":2592,"would_cite":false,"duration_ms":26470,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["76N10","35B45"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes global classical solutions and optimal decay rates for the 3D isentropic compressible Navier–Stokes equations with density-dependent viscosities when initial perturbations are small in L1∩L2 and α is close to 1…","keywords":["compressible Navier-Stokes equations","density-dependent viscosity","global well-posedness","optimal decay rates","Green's function method","regularity criterion","Cauchy problem","classical solutions"],"falsifier":"Take initial data V_0=(ϱ_0,u_0) as a bump of height M and width w with M $w^{{3/2}}$=C_0: the $L^{2}$ norm is C_0 and the $L^{1}$ norm is $C_0^{2}$/M, so for large M the data satisfy (1.7) while violating the pointwise bound at t=0. A check of the linear estimate (3.42) or (3.51) with the factor ∥∇^2V_0∥^{3/4} kept would settle whether the claimed decay holds without an additional high-frequency assumption.","tokens_in":20745,"feed_emoji":"🌊","tokens_out":10390,"duration_ms":99154,"temperature":0.7,"pith_summary":"The paper claims that the three-dimensional isentropic compressible Navier–Stokes equations with density-dependent viscosities μ=λ=ρ^α admit a unique global classical solution whenever the initial perturbation of a constant density state is small in $L^{1}$∩$L^{2}$ and the viscosity exponent satisfies |α-1|≤δ, no matter how large the $H^{4}$ Sobolev norms of spatial derivatives of the data are. The solution is shown to decay like a heat kernel, with ∥(ϱ,u)∥_{L^∞}≤C(1+t)^{-3/2}, ∥∇(ϱ,u)∥_{L^∞}≤C(1+t)^{-2}, and ∥∇^k(ϱ,u)∥_{$L^{2}$}≤C(1+t)^{-3/4-k/2} for k=0,1,2, together with matching L^p rates. This sharpens earlier decay results that required smallness of the whole Sobolev norm of the initial data; if correct, it means the long-time behavior is controlled by the low-frequency, integrable part of the data alone, while high-frequency oscillations may be large. The proof combines a Green's function decomposition of the linearized flow with a time-decay regularity criterion and conditional energy estimates, and the near-constant-viscosity condition α≈1 is used to absorb higher-order dissipation terms.","feed_headline":"Global decay for compressible flows from small L1/L2 data","feed_subtitle":"No smallness of Sobolev derivatives needed; only the viscosity exponent near 1 is required.","key_machinery":"The load-bearing object is the Green function G(t,x) of the linearized system, decomposed into a low-frequency heat-like part G_L, a high-frequency regular part G_HR that can bear one derivative and decays exponentially, and a singular high-frequency part G_HS that behaves like $e^{{-ct}}$δ(x) and appears only in the continuity equation. The argument runs through the Duhamel formula V(t)=G(t)*V_0+∫_0^t G(t-s)*N(V)(s)ds, and the bootstrap is a time-decay regularity criterion: assuming ∥(ϱ,u)∥_{L^∞}≤1/2(1+t)^{-3/2} and ∥∇(ϱ,u)∥_{L^∞}≤η(1+t)^{-2}, the paper proves conditional energy estimates needing only $L^{2}$ smallness, then closes the criterion with an auxiliary $L^{{4/3}}$ estimate and the Green function's L^p decay. The viscosity constraint |α-1|≤δ enters through H_α(ϱ)=(1+ϱ)^{α-1}-1, whose L^∞ norm is O(|α-1|), allowing the high-order velocity dissipation terms in Lemma 3.4 to be absorbed.","core_discovery":"On its own terms, the central discovery is Theorem 1.1: with (ϱ_0,u_0)∈$H^{4}$, ∥(ϱ_0,u_0)∥_{$L^{1}$∩$L^{2}$}≤C_0 and |α-1|≤δ, the Cauchy problem has a unique global classical solution (ϱ,u)∈C([0,∞);$H^{4}$) with ∇u∈$L^{2}$([0,∞);$H^{4}$), and the optimal uniform decay (1.9)–(1.11) holds. The point of the theorem is that no smallness is imposed on any Sobolev norm of derivatives; the $H^{4}$ norm of the initial data may be arbitrarily large. The mechanism is a bootstrap that starts from the time-decay regularity criterion (3.1), proves conditional energy estimates that need only $L^{2}$ smallness, and then closes the criterion through the Duhamel representation using low-frequency heat-like decay of the Green function and exponential high-frequency estimates. The condition |α-1|≤δ is essential to the closure, since the linearized viscosity perturbation H_α(ϱ)=(1+ϱ)^{α-1}-1 is of size O(|α-1|) and can be absorbed into the velocity dissipation.","pith_inferences":["An extension the authors do not pursue: the auxiliary L^{4/3} estimate in Lemma 3.5 could likely be replaced by any L^q with q<2, since it only supplies a small algebraic decay surplus; the final rates would be unchanged and the interpolation might simplify.","If the mechanism is robust, the same Green-function-plus-regularity-criterion bootstrap should transfer to other systems whose linearization has a heat-like low-frequency block, such as viscous shallow water equations, giving analogous no-small-derivative decay results.","Because the bootstrap requires the criterion (3.1) at t=0, the advertised regime of arbitrarily large derivatives is fully valid only if a short-time estimate can force the initial pointwise bounds; supplying such an estimate would complete the theorem exactly as stated."],"forward_implications":["Global classical solutions exist for all time for small L^1∩L^2 perturbations of a constant-density state even when the H^4 norms of initial derivatives are large.","The density and velocity approach the equilibrium at heat-like rates: ∥(ϱ,u)∥_{L^∞}≤C(1+t)^{-3/2} and ∥∇(ϱ,u)∥_{L^∞}≤C(1+t)^{-2}.","The H^k norms for k=0,1,2 decay at the rates ∥∇^k(ϱ,u)∥_{L^2}≤C(1+t)^{-3/4-k/2}.","The L^p interpolation bounds (1.11) hold for 2≤q≤∞ with k'=0,1.","The theorem also gives uniqueness of the global classical solution, so the asymptotic state is determined by the low-frequency part of the data even when high-frequency derivatives are large."],"supporting_citations":[{"why":"Supplies the energy-decay method for dissipative systems whose whole Sobolev-norm smallness assumption is removed here.","marker":"[4]"},{"why":"Supplies Green's function and diffusion-wave decay estimates for compressible flow used in the linear analysis.","marker":"[7]"},{"why":"Supplies pointwise decay estimates for the multidimensional Navier–Stokes Green function, including high-frequency behavior.","marker":"[8]"},{"why":"The prior density-dependent viscosity decay result with small derivatives, which the paper's partial smallness condition improves.","marker":"[10]"},{"why":"Provides Green-function regularity estimates for the viscous shallow water system used to justify Lemma 2.2.","marker":"[16]"},{"why":"Gives the local existence of classical solutions used to start and extend the continuity argument.","marker":"[20]"},{"why":"Supplies global well-posedness with highly oscillating initial velocity, one of the Green-function estimates invoked in the paper.","marker":"[3]"}],"fun_headline_variants":["No Sobolev smallness needed for compressible flow decay","Viscosity near 1 gives optimal decay with large Sobolev norms","Compact: global decay from L1∩L2 smallness, α≈1","Compressible Navier-Stokes: decay without Sobolev smallness"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof's bootstrap needs the initial data to satisfy the pointwise bounds ∥(ϱ_0,u_0)∥_{L^∞}≤1/2 and ∥∇(ϱ_0,u_0)∥_{L^∞}≤η, but smallness of the $L^{1}$∩$L^{2}$ norm alone does not force those bounds when higher-order derivatives are large.","fun_headline_variants_meta":{"raw":{"variants":["No Sobolev smallness needed for compressible flow decay","Viscosity near 1 gives optimal decay with large Sobolev norms","Compact: global decay from L1∩L2 smallness, α≈1","Compressible Navier-Stokes: decay without Sobolev smallness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000199,"raw_usage":{"total_tokens":1371,"prompt_tokens":946,"completion_tokens":425,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":562,"completion_tokens_details":{"reasoning_tokens":344}},"tokens_in":562,"tokens_out":425,"duration_ms":4824,"temperature":1.0,"reasoning_tokens":344,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:10:23.338562+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take initial data V_0=(ϱ_0,u_0) as a bump of height M and width w with M $w^{{3/2}}$=C_0: the $L^{2}$ norm is C_0 and the $L^{1}$ norm is $C_0^{2}$/M, so for large M the data satisfy (1.7) while violating the pointwise bound at t=0. A check of the linear estimate (3.42) or (3.51) with the factor ∥∇^2V_0∥^{3/4} kept would settle whether the claimed decay holds without an additional high-frequency assumption.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the energy-decay method for dissipative systems whose whole Sobolev-norm smallness assumption is removed here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Green's function and diffusion-wave decay estimates for compressible flow used in the linear analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies pointwise decay estimates for the multidimensional Navier–Stokes Green function, including high-frequency behavior."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The prior density-dependent viscosity decay result with small derivatives, which the paper's partial smallness condition improves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides Green-function regularity estimates for the viscous shallow water system used to justify Lemma 2.2."},{"cited_title":"Zhang, J","cited_arxiv_id":null,"evidence_quote":"Gives the local existence of classical solutions used to start and extend the continuity argument."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies global well-posedness with highly oscillating initial velocity, one of the Green-function estimates invoked in the paper."}],"review_version":1}