{"id":"6aa05df9-5185-4919-b964-6f67142671c1","arxiv_id":"2607.13567","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"high","formal_verification":"none","parameter_count":2,"one_line_summary":"Constructive computer-assisted existence of overhanging periodic gravity water waves on a global branch is claimed, with termination by self-intersection at the trough line; the written proof of the key injectivity lemma has a sign error.","lead":"This paper reports a computer-assisted proof that periodic water waves with strong vorticity bend into overhanging shapes along a global branch from flat water, ending by self-intersection at the trough. A generalist should care because it is the first claimed rigorous proof of such overhanging waves at fixed macroscopic parameters—though a key geometric lemma has an invalid sign step.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 4.2's exponential tail bound is false: for a supported at −δ, the commutator residual coefficient is ≈ −2δ, not O(e^{-2hK0}), so the CAP tail bounds and Theorems 8.3–8.4 are unproven.","rationale":"The paper contains substantial constructive material: explicit code, interval arithmetic, and a detailed continuation framework. However, the central proof relies on a false analytic lemma. Lemma 4.2 is not a minor technical estimate; it is the mechanism that makes the infinite-dimensional tail of the Newton–Kantorovich inverse exponentially small. If the residual is instead O(K0)‖a‖_X, the stated Y, Z1, Z2 bounds—and hence the verified radii polynomials in Table 1—are not rigorously justified. The main existence and global-branch theorems therefore are not proven as written. I do not agree with the reader's diagnosis: the Hopf sign in Lemma 2.2 is actually correct, because the conformal domain is (0,π)×(−h,0) and the outward normal at y=0 is +∂_y. The load-bearing flaw lies in Lemma 4.2's cancellation argument. A corrected proof might be possible—the residual being O(K0) is still finite and could be incorporated—but the current manuscript should be rejected or majorly revised until the commutator estimate is fixed and the continuation bounds are recomputed.","tokens_in":59761,"tokens_out":10881,"duration_ms":108779,"concrete_test":"Compute the single-mode commutator residual exactly. Let h=2, K0=200, K1=402, choose a with a_{−1}=1, and b=δ_n with n=K1+1. The coefficient at k=n−1 is ω_{n−1}−ω_n−1 = (401)coth(802)−(403)coth(806)−1 ≈ −2. The RHS of Lemma 4.2 is < 1e−300. This one-mode evaluation falsifies the claimed exponential bound. Equivalently, re-derive the Σ_int term without the false identity |k|−|n|=|k−n|; the exact residual is O(K0)‖a‖_X, and recomputing Z∞ with that term will show whether Table 1's radii polynomials still hold.","verdict_should_be":"REJECT","load_bearing_attack":"Lemma 4.2 claims that C[a]bπ_{>K1} − M_{D_a}bπ_{>K1} is exponentially small, with bound 4K0(1−e^{−2hK0})^{-1}e^{−2hK0}‖a‖_X. The proof splits into Σ_res and Σ_int and asserts Σ_int = 0 by cancellation. This is incorrect. Take a supported at mode −δ (|δ| ≤ K0) and an input b at frequency n > K1. The output is at k = n − δ, and the coefficient of C[a] − M_{D_a} is ω_{n−δ} − ω_n − |δ|. For large n, ω_{n−δ} ≈ n−δ and ω_n ≈ n, so the coefficient is approximately −2δ, not zero. The proof's cancellation mistakenly replaces |k| − |n| by |k−n|: with k = n−δ, |k|−|n| = −δ while |k−n| = δ. Thus the residual is not exponentially small; it contains terms of size 2δ for every high-frequency input. Lemmas 4.6, 5.2, and 6.3 rely on Lemma 4.2 to control the commutator in the Z∞ bound, hence in Z1. The rigorous Newton–Kantorovich bounds and the computer-assisted continuation proofs in Theorems 8.3 and 8.4 are therefore not established as written. The reader's objection to Lemma 2.2, by contrast, does not land: at y=0 the outward normal is +∂_y, so Hopf gives ξ_y<0 and hence η_x>0, a genuine contradiction.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a conformal Fourier formulation for steady two-dimensional periodic gravity water waves with constant vorticity, allowing multi-valued surface heights. It combines qualitative maximum-principle and Hopf-boundary-lemma arguments with a computer-assisted Newton–Kantorovich continuation framework. The main theorems claim: (i) existence of a unique overhanging 'Omega-shaped' exact wave at m=-0.85, and (ii) a smooth global branch from the flat state at m≈-0.061 down to m=-1 that transitions from graphs to overhanging waves and terminates by self-intersection at the trough line. The computer-assisted part uses explicit truncations K0=200, K1=402 and reports rigorous bounds Y, Z1, Z2 (Table 1), with the stated goal of partially resolving the Constantin–Strauss–Varvaruca conjecture.","tokens_in":60199,"tokens_out":14364,"duration_ms":145682,"significance":"If the result were correct, it would be a significant advance: the first rigorous construction of overhanging periodic gravity water waves along a finite-depth global bifurcation branch from flat water at fixed O(1) parameters, with a concrete termination mechanism. The paper's strengths include a clear functional-analytic setup, explicit and reproducible computational artifacts (code available at [10]), and the use of standard Newton–Kantorovich machinery with stated constants. Unfortunately, a central analytic estimate in the computer-assisted framework is false, so the main existence and continuation theorems are not established as written.","major_comments":[{"comment":"Lemma 4.2 is false. In the proof, after the finite-depth discrepancy is collected into Sigma_res, the term Sigma_int is declared to vanish using the identity |k|-|n|=|k-n|. The threshold K1>=2K0 only gives sgn(k)=sgn(n), which yields |k|-|n|=sgn(n)(k-n), not |k-n|. For k=n-delta with 0<delta<=K0, |k|-|n|=-delta while |k-n|=delta. Concretely, take a with a_{-1}=1 and u with u_n=1 for n>K1. The (n-1)-coefficient of (C[a]pi_{>K1}-M_Da pi_{>K1})u is omega_{n-1}-omega_n-omega_{-1}-D_{-1}, which tends to -2, not O(e^{-2hK0}). Thus the high-frequency residual contains O(1) terms for every large n, contradicting the exponential bound claimed in the lemma.","section":"Lemma 4.2 (Section 4.3)"},{"comment":"The false bound in Lemma 4.2 is load-bearing for the entire computer-assisted argument. Lemma 4.6 uses Lemma 4.2 to control the commutator contribution to Z_infty, and Lemmas 5.2 and 6.3 are the continuation versions of the same estimate. Consequently, the reported Z_1 values in Theorem 8.1 and Table 1 (e.g., Z_1=5.78e-3 on [-0.2,-0.1]) are not justified. The contraction inequalities (4.4) and (6.2) have therefore not been verified, and Theorems 8.1, 8.3, and 8.4 do not follow from the numerical data. This is not a matter of tightening constants: the missing terms are O(1), not O(e^{-2hK0}).","section":"Lemmas 4.6, 6.3; Table 1; Theorems 8.1, 8.3, 8.4"}],"minor_comments":[{"comment":"The stress-test concern about the Hopf sign in this lemma does not land. At the top boundary y=0 the outward normal for the domain -h<y<0 is +partial_y. Hopf gives xi_y<0, and the Cauchy-Riemann relation xi_y=-eta_x then gives eta_x>0, exactly the contradiction stated in the paper.","section":"Lemma 2.2 (Section 2.2)"},{"comment":"The abstract says the analysis 'resolves the conjecture' of [17], while Section 1 says 'partially resolving.' Since the theorem concerns one specific branch in one parameter regime, the weaker wording is more accurate and should be used consistently.","section":"Abstract and Section 1.3"},{"comment":"The admissible set is defined as an open set in the C^2 topology, but the continuation framework works in the H^1-type space X (weighted l^1 controlling one derivative). The paper should justify that the constructed branch is continuous in a topology compatible with Lemma 2.3, or give a separate openness argument adapted to the spaces actually used.","section":"Lemma 2.3 and Corollary 2.4"}],"recommendation":"reject","confidential_remarks":"The paper is ambitious, well organized, and the computational infrastructure appears reproducible. However, the central analytic commutator estimate in Lemma 4.2 is false, and the reported Newton-Kantorovich bounds therefore do not certify the claimed existence or continuation results. Because the error is O(1) rather than exponentially small, repairing it is not a local fix; substantial reworking of the tail analysis would be needed. I would encourage the authors to address this point and resubmit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: don't trust the main theorems yet. The reader's grounds are wrong—Lemma 2.2 is fine—but the stress-test note is right: Lemma 4.2's exponential tail bound is false, and everything downstream inherits the problem.\n\nThe genuinely new piece is the goal itself: first rigorous attack on overhanging periodic waves along a finite-depth global branch at fixed O(1) parameters, using a conformal-mapping formulation plus computer-assisted interval arithmetic. The qualitative part (monotonicity, admissible open sets, gluing arguments) is careful and mostly correct. The public code and explicit bounds are real work product of the right kind.\n\nOn the specific objections: at the top boundary y=0 the outward normal is +∂_y, so a boundary minimum gives ξ_y<0, and Cauchy–Riemann gives η_x>0, which contradicts strict monotonicity. Lemma 2.2 stands. The reader's sign flip is not there.\n\nThe real problem is Lemma 4.2. The proof splits the commutator residual and asserts Σ_int=0 by the identity |k|−|n|=sgn(n)(k−n). That only cancels when (k−n) has the same sign as n and therefore equals |k−n|; when k−n is opposite in sign, the difference is 2(k−n), not zero. Concretely, for a supported near mode −δ with δ>0, a high-frequency input at n>K1 produces an output at k=n−δ whose coefficient is roughly −2δ, not O(e^{−2hK0}). So C[a] is not exponentially close to M_{D a} on the tail. Since Lemma 4.6 (and its parameter-dependent versions 5.2 and 6.3) use exactly this to bound Z∞, hence Z1, the Newton–Kantorovich radii in Theorems 8.1–8.4 are not justified. This is load-bearing, not a typo.\n\nI'd also note the abstract says \"resolves\" while the body says \"partially resolves\"; minor, but it should be aligned.\n\nWho should read it: people working on CAP for water waves will want to see the framework and maybe repair it. It deserves a serious referee, because the failure is local and the target is real. But as written it should not be accepted; the authors need to replace Lemma 4.2 and recompute the bounds.","headline":"Main theorems are unproven because Lemma 4.2's cancellation is false—not for the reason the reader gave; the paper is a serious but currently invalid CAP contribution.","tokens_in":60664,"tokens_out":8296,"would_cite":false,"duration_ms":74526,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B32","35Q35","76B15"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims a constructive proof that a constant-vorticity periodic water wave branch from flat state transitions to overhanging profiles and terminates by self-intersection at the trough line.","keywords":["overhanging water waves","global bifurcation","constant vorticity","computer-assisted proof","Newton-Kantorovich","conformal mapping","periodic gravity waves","self-intersection"],"falsifier":"Check the sign in the crest-axis lemma: at a boundary minimum of ξ on y=0, the outward normal is -∂_y, so Hopf gives -ξ_y<0, i.e. ξ_y>0, and by Cauchy-Riemann η_x<0, which is consistent with strict monotonicity rather than contradicting it. Computationally, rerun the interval-arithmetic geometric verification for m∈[-0.9,-0.85] while monitoring min_x ξ(x,0) on (0,π): if it reaches zero before the branch enters the claimed self-intersection interval, crest-axis crossing is the actual termination mechanism.","tokens_in":1647,"feed_emoji":"🌊","tokens_out":2619,"duration_ms":85924,"temperature":0.7,"pith_summary":"The paper aims to prove that overhanging periodic gravity water waves with constant vorticity exist as exact solutions reachable by continuing a branch from flat water at fixed macroscopic parameters (gravity 1, depth 2, vorticity -5). It develops a conformal mapping formulation that allows multi-valued surface heights, reducing the problem to a zero-finding equation on Fourier coefficients, and uses a computer-assisted Newton-Kantorovich argument to rigorously enclose the entire branch. The main payoff is the first rigorous proof that a finite-depth global branch from flat water turns over and that its topological termination is a self-intersection across the trough line. The central qualitative claim is that a strictly monotone wave cannot cross the crest axis, so only trough-line self-intersection needs to be monitored.","feed_headline":"Overhanging periodic waves proven along global branch from flat water","feed_subtitle":"First rigorous proof that constant-vorticity water waves turn over and end in trough-line self-intersection at fixed order-one parameters.","key_machinery":"The central object is a conformal map X+iY=ξ(x,y)+iη(x,y) from a rectangle onto the fluid domain, turning the free-boundary problem into a system of harmonic functions with nonlinear boundary conditions. Fourier reduction with the Dirichlet-to-Neumann multiplier ω_n=|n|coth(|n|h) reduces the problem to a one-dimensional zero-finding equation F(U)=0 on even Fourier sequences. The global existence proof uses a uniform Newton-Kantorovich theorem with an approximate inverse split into a finite-dimensional Galerkin block and an analytic high-frequency tail. A desingularized scaling handles the pitchfork bifurcation, and Chebyshev interpolation in m provides uniform bounds. Geometric classificatio","core_discovery":"The central claim is the constructive existence of a C-infinity branch of exact 2π-periodic gravity water waves with constant vorticity γ=-5, depth h=2, and gravity g=1, parameterized by mass flux m∈[-1,m], with m≈-0.061. The branch bifurcates from flat water, consists of graphs near the bifurcation point, becomes strictly overhanging below a first critical flux in [-0.698,-0.6707], and loses physical injectivity below a second critical flux in [-0.8884,-0.8781], where the profile self-intersects across the trough line. This partially resolves the Constantin-Strauss-Varvaruca conjecture on branch termination and provides the first rigorous overhanging periodic waves reached from flat water a","pith_inferences":["Editorial inference: if the crest-axis injectivity lemma is repaired or replaced, the same continuation strategy could map, across vorticity and depth parameters, which regimes end in overhanging-touching waves versus waves of greatest height.","Editorial inference: the proof certifies physical validity only against trough-line self-intersection; a cheap testable extension would add a rigorous lower bound on ξ(x,0) throughout the continuation, which the existing interval-arithmetic machinery could do.","Editorial inference: the conformal Fourier formulation with rigorous continuation may transfer to solitary or internal overhanging waves, where domain-flattening methods fail at the overhang.","Editorial inference: because the validated branch is smooth in m, its rigorous Fourier coefficients could support computer-assisted stability or spectral computations for these overhanging profiles."],"forward_implications":["Exact, smooth, stagnation-free overhanging periodic waves exist at fixed order-one physical parameters with constant vorticity, not only in perturbative small-gravity regimes.","In this parameter regime the global branch terminates by self-intersection on the trough line, so the limiting behavior is a touching-type degeneration rather than a Stokes corner.","The branch is infinitely differentiable in the mass-flux parameter m across the whole continuation, including the transition from graphs to overhangs.","The framework yields explicit rigorous enclosures of the solutions and of the critical fluxes marking the onset of overhanging and of self-intersection.","For flux values below the second critical value the mathematical branch continues to exist but no longer describes a physically valid single-valued wave."],"fun_headline_variants":["First rigorous proof of overhanging periodic water waves","Overhanging waves: first rigorous proof from flat water","Global branch proves overhanging periodic waves exist","Conjecture on overhanging waves confirmed via global branch"],"cache_read_input_tokens":61824,"weakest_assumption_plain":"Everything downstream depends on the claim that a strictly monotone wave cannot cross the crest axis, i.e. ξ(x,0)>0 on (0,π); as written, the proof of that lemma applies the Hopf boundary lemma with the wrong normal direction at the free surface, so the lemma is not established in the text, and if it fails the branch could lose physical validity by a crest-axis crossing before the claimed trough-line self-intersection.","fun_headline_variants_meta":{"raw":{"variants":["First rigorous proof of overhanging periodic water waves","Overhanging waves: first rigorous proof from flat water","Global branch proves overhanging periodic waves exist","Conjecture on overhanging waves confirmed via global branch"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000945,"raw_usage":{"total_tokens":3865,"prompt_tokens":729,"completion_tokens":3136,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":473,"completion_tokens_details":{"reasoning_tokens":3080}},"tokens_in":473,"tokens_out":3136,"duration_ms":21689,"temperature":1.0,"reasoning_tokens":3080,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T04:47:22.080065+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the sign in the crest-axis lemma: at a boundary minimum of ξ on y=0, the outward normal is -∂_y, so Hopf gives -ξ_y<0, i.e. ξ_y>0, and by Cauchy-Riemann η_x<0, which is consistent with strict monotonicity rather than contradicting it. Computationally, rerun the interval-arithmetic geometric verification for m∈[-0.9,-0.85] while monitoring min_x ξ(x,0) on (0,π): if it reaches zero before the branch enters the claimed self-intersection interval, crest-axis crossing is the actual termination mechanism.","supporting_citations":[],"review_version":1}