{"id":"7432b7f8-a05e-447c-abd7-a067d442dfc6","arxiv_id":"2501.01390","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For p=2,3,4, the coefficient beta_1^{(p)}(h) controlling the exponentially small width of the p-th Stokes-wave isolas has explicit deep-water asymptotics, namely (3*sqrt(3)/64)e^{-h/2}, (2*sqrt(2)/3)e^{-2h}, and -(5*sqrt(5)/(8*sqrt(3)))e^{-2h}.","lead":"Stokes waves are periodic traveling water waves that show small high-frequency instabilities, called isolas. This paper computes exactly how fast the first three such instabilities shrink as the ocean becomes infinitely deep.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The p=4 branch of Theorem 2.1 hinges on the unverified Mathematica cancellation (2.56); a nonzero e^{-2h} term there would change the leading coefficient in (2.3).","rationale":"The reader's weakest_assumption identifies precisely the p=4 cancellation (2.56) as the fragile spot, and I agree. The paper's central claim has three independent cases; p=2 and p=3 are derived analytically and are solid. For p=4, the leading coefficient is obtained by summing eight explicit groups, and the eight-term group (2.38s)-(2.38z) is dismissed with a single Mathematica-computed O(e^{-4h}) statement. Because every term in that group has an e^{-2h} part, the cancellation is not forced by the known deep-water limit (which only cancels the order-1 constants); it is an additional algebraic identity that must be checked. The approximate numerical sum of the explicit e^{-2h} coefficients does match the claimed -5√5/(8√3) if (2.56) is zero, which gives some internal consistency, but it does not verify the cancellation. Without a reproducible computation or a written derivation, the p=4 branch is unverified, so the theorem as stated should be accepted only conditionally on that check. This is a modest adjustment, not a rejection: the methodology is sound, and the concern is about verifiability, not a detected falsehood.","tokens_in":22101,"tokens_out":11836,"duration_ms":101810,"concrete_test":"Independently expand the eight expressions (2.38s)-(2.38z) to order e^{-2h} using the stated expansions (2.9), (2.10), (2.24), (2.41), and (2.44), and compute the signed combination in (2.56); verify that the coefficient of e^{-2h} is exactly zero and the remainder is O(e^{-4h}). A direct alternative is to run the linked Mathematica notebook and confirm it produces exactly this cancellation. If the notebook is inaccessible or the cancellation fails, the p=4 part of Theorem 2.1 is unsupported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem 2.1 for p=4 reduces to a signed sum of 27 terms (2.37). All groups except the last are expanded explicitly to O(e^{-4h}), but the eight three-intermediate-harmonic terms in (2.38s)-(2.38z) are combined in (2.56) and asserted to sum to O(e^{-4h}) \"by a very long computation using Mathematica\". Each of those eight terms is of order 1 and has an e^{-2h} correction (via (2.41) and (2.44)), so the cancellation of both the order-1 and the e^{-2h} parts is nontrivial. If the e^{-2h} coefficient of the signed sum in (2.56) were c≠0, then (2.3) would become (-5√5/(8√3)+c)e^{-2h}+O(e^{-4h}), invalidating the p=4 claim and the finite-zeros/Conjecture 2.2 consequence for p=4. The paper gives no written derivation or intermediate output; the linked notebooks are the only check, and the text does not document a commit hash or a standalone verification script. The p=2 and p=3 branches are analytic and are not affected by this concern.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the coefficient beta_1^{(p)}(h) that controls the width of the p-th high-frequency instability isola of Stokes waves in the deep-water limit. Building on the formula for beta_1^{(p)} as a sum of 3, 9, and 27 explicit terms for p=2,3,4 respectively, it proves Theorem 2.1, giving the sharp exponential decay rates beta_1^{(2)}~3*sqrt(3)/64*e^{-h/2}, beta_1^{(3)}~2*sqrt(2)/3*e^{-2h}, and beta_1^{(4)}~-5*sqrt(5)/(8*sqrt(3))*e^{-2h} as h->+infty. The p=2 and p=3 cases are derived analytically in the text, while the p=4 case relies on a Mathematica-assisted cancellation among the eight three-intermediate-harmonic terms. The paper also concludes that these beta_1^{(p)} have finitely many positive zeros, supporting Conjecture 2.2.","tokens_in":22345,"tokens_out":13890,"duration_ms":113681,"significance":"If the p=4 computation can be verified, the result provides sharp asymptotic widths for the p=2,3,4 isolas in deep water, confirms the exponential vanishing of beta_1^{(p)} at infinity, and establishes finiteness of zeros for these p. The p=2 and p=3 derivations are transparent and reproducible from the displayed formulas; the paper is also honest about the computational assistance used for p=4 and the analytic openness of p>4. The result is a meaningful complement to the existence theory in the authors' prior work [3] and lends support to a natural conjecture. The main caveat is that the p=4 branch rests on an unverified cancellation that is not documented in the text.","major_comments":[{"comment":"The assertion that the eight three-intermediate-harmonic terms sum to O(e^{-4h}) is load-bearing for the p=4 case of Theorem 2.1, but no derivation or intermediate output is given in the text; the only justification is 'a very long computation using Mathematica'. Since each individual term is of order one and has an e^{-2h} correction (via (2.41) and (2.44)), the cancellation of both the order-one and e^{-2h} parts is nontrivial. If the e^{-2h} coefficient of the signed sum were a nonzero c, formula (2.3) would become (-5*sqrt(5)/(8*sqrt(3))+c)e^{-2h}+O(e^{-4h}), changing the leading coefficient and the finiteness-of-zeros conclusion for p=4. Please provide the explicit expanded form of the eight terms to order e^{-4h}, or include a reproducible script with output, so that (2.56) can be checked without trusting an opaque computation.","section":"Section 2.3, Eq. (2.56)"},{"comment":"The two displayed expansions for [b(4)_1]^-_2 and [b(4)_1]^+_2 are printed identically, despite the different signs in their definitions (2.38c)-(2.38d). If this is a typo and the e^{-2h} coefficients differ, then (2.49) would not be O(e^{-4h}) and the leading coefficient in (2.3) would be affected. The authors should either verify that the two coefficients are indeed equal (explaining the cancellation) or correct the typo and rerun the computation.","section":"Section 2.3, Eq. (2.48)"}],"minor_comments":[{"comment":"The name 'Zhakarov' should be 'Zakharov'.","section":"Section 1, page 1"},{"comment":"Reference [12] (Creedon and Deconinck) is listed in the bibliography but never cited in the text.","section":"References"},{"comment":"The proof of Lemma 2.5 is only a reference to [3, formulas (A.11),(2.31)]; for self-containedness, a short derivation analogous to Lemmas 2.3-2.4 would help, since this expansion is a load-bearing input for the p=4 computation.","section":"Section 2.3, Lemma 2.5"},{"comment":"If the identical expansions in (2.48) are intentional, a note explaining the cancellation would aid the reader; otherwise the identical appearance is a likely typo. More generally, a table listing the individual contributions of the 27 terms would improve readability.","section":"Section 2.3, Eq. (2.56)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript depends heavily on the authors' own preprint [3] (arXiv:2405.05854), which does not yet appear to be published. The p=4 branch also relies on an unverified Mathematica cancellation in (2.56). If [3] is not yet accepted, the editors may wish to confirm its availability and correctness. The p=2 and p=3 results are solid and fully analytic; the p=4 gap is fixable by adding the missing computation or a reproducible script, so the paper is not beyond repair."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know about this paper is that it does exactly what it says: for p=2,3 it gives a hands-on analytic derivation of the exponential decay rates in Theorem 2.1, and for p=4 it gives a plausible computer-assisted derivation. That is more than the previous paper [3] had, which only proved the limit to zero. The result is the leading constants and rates, and it settles the finite-zeros part of Conjecture 2.2 for these three cases. I think the significance is real but modest—it sharpens the picture of high-frequency instability isolas without changing the qualitative story.\n\nWhat I like: the p=2 and p=3 sections are fully checkable by hand. I spot-checked several of the expansions (2.17)–(2.19) and (2.30)–(2.36); the arithmetic lines up. The grouping by number of intermediate harmonics is a nice way to keep 27 terms organized. The paper is honest that p=4 is computational and points to notebooks.\n\nThe soft spot is exactly where the stress-test note lands. The eight three-intermediate-harmonic terms in (2.38s)–(2.38z) are each order one, and each carries an e^{-2h} correction via (2.41) and (2.44). In (2.56) their signed sum is asserted to be O(e^{-4h}) by 'a very long computation using Mathematica'. That is a heavy cancellation, and it is load-bearing: if the e^{-2h} coefficient in that sum were nonzero, the leading constant in (2.3) would change. The notebooks are a good-faith attempt, but without a commit hash or a minimal script that prints the intermediate grouped sums, a referee can't verify (2.56) without rerunning the whole notebook and trusting no transcription error. I would not call this fatal—the p=2,3 results stand alone, and the p=4 coefficient is likely correct—but it deserves to be flagged as the one fragile point.\n\nAlso minor: the paper quotes the formulas for beta_1^(p) from [3] rather than rederiving them. That is fine as a companion piece, but it means the full proof inherits the correctness of [3], which is a much longer paper.\n\nOverall: this is a clean, useful companion paper. The audience is people working on spectral instability of Stokes waves; they'll get a precise asymptotic reference. I'd send it out. I'd ask the authors to add an appendix or notebook output that verifies the individual e^{-2h} coefficients in the eight terms of (2.56) before it is final.","headline":"Sharp deep-water asymptotics for the first three Stokes-wave isolas; p=2,3 are clean, p=4 hangs on a Mathematica cancellation that should be independently checkable.","tokens_in":22910,"tokens_out":2277,"would_cite":true,"duration_ms":21685,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","35B35","76B15"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper computes sharp deep-water asymptotics for the coefficients $\\beta_1^{(2)}$, $\\beta_1^{(3)}$, $\\beta_1^{(4)}$ that size the high-frequency instability isolas of Stokes waves, showing they decay exponentially and vanish only…","keywords":["Stokes waves","high-frequency instability","isolas","modulational instability","deep water limit","asymptotic expansion","water waves","spectral theory"],"falsifier":"Recompute the signed sum in (2.56) with independent symbolic or high-precision arithmetic: if the eight three-intermediate-harmonic terms do not cancel to $O(e^{-4h})$, the $p=4$ statement falls. Alternatively, evaluate $\\beta_1^{(4)}(h)$ numerically at large $h$ (for example $h=20$) using the formulas of Section 2.3 and compare with $-\\frac{5\\sqrt{5}}{8\\sqrt{3}}e^{-2h}$; a mismatch beyond the stated $O(e^{-4h})$ error would refute Theorem 2.1.","tokens_in":21897,"feed_emoji":"🌊","tokens_out":15808,"duration_ms":132299,"temperature":0.7,"pith_summary":"Stokes waves are periodic traveling water waves, and a central question is whether small perturbations grow. The linearized problem has a sequence of closed spectral bands—isolas—away from the origin, indexed by $p\\geq 2$, whose existence and size are controlled by an analytic function $\\beta_1^{(p)}(h)$ of the water depth $h$. This paper proves that as $h\\to+\\infty$ the coefficients for $p=2,3,4$ decay exponentially, with the sharp leading terms $\\beta_1^{(2)}\\sim \\frac{3\\sqrt{3}}{64}e^{-h/2}$, $\\beta_1^{(3)}\\sim \\frac{2\\sqrt{2}}{3}e^{-2h}$, and $\\beta_1^{(4)}\\sim -\\frac{5\\sqrt{5}}{8\\sqrt{3}}e^{-2h}$. Because the p-th isola has width proportional to $|\\beta_1^{(p)}|\\epsilon^p$, this quantifies how the first three high-frequency instability bands shrink in deep water. It also gives an independent proof that $\\beta_1^{(p)}$ is not identically zero for $p=2,3,4$ and has only finitely many depth zeros, supporting a conjecture that the same is true for every $p$.","feed_headline":"Exact decay rates found for the first three Stokes instability bands","feed_subtitle":"In deep water the p=2,3,4 instability bands shrink at explicit exponential rates, proving they persist at all sufficiently large depths.","key_machinery":"The central object is the coefficient $\\beta_1^{(p)}(h)$: the leading term in the discriminant $D^{(p)}(\\mu,\\epsilon)$ of the two eigenvalues of the Floquet operator $L_{\\mu,\\epsilon}$ near the double eigenvalue $i\\omega_*^{(p)}(h)$ at $\\mu=\\varphi(p,h)$. When it is nonzero, the sign of $D^{(p)}$ changes across an interval of Floquet exponents of width proportional to $|\\beta_1^{(p)}|\\epsilon^p$, producing the isola. The paper computes its deep-water limit from explicit formulas in which $\\beta_1^{(p)}$ is a sum of rational products of Stokes-wave Taylor-Fourier coefficients ($a_\\ell^{[\\ell]}$, $p_\\ell^{[\\ell]}$) and the quantities $\\Omega_j^{(p)}$, $t_j^{(p)}$ built from the wavenumber $\\varphi(p,h)$; Lemmas 2.3–2.5 refine the expansion of $\\varphi(p,h)$ as $h\\to\\infty$, and the leading exponentials emerge after cancellations of the constant parts. For $p=4$ the cancellation of the eight three-intermediate-harmonic terms (equation (2.56)) is checked by a long computer-algebra computation.","core_discovery":"On the paper's own terms, the central result is Theorem 2.1: in the deep-water limit $h\\to+\\infty$, the coefficient $\\beta_1^{(p)}(h)$ that enters the discriminant of the two relevant eigenvalues of the Floquet operator has the asymptotic expansions $$\\$beta_1^{{(2)}}$(h)=\\frac{3\\sqrt{3}}{64}$e^{{-h/2}}$+O($e^{{-3h/4}}$),\\qquad \\$beta_1^{{(3)}}$(h)=\\frac{2\\sqrt{2}}{3}$e^{{-2h}}$+O($e^{{-3h}}$),\\qquad \\$beta_1^{{(4)}}$(h)=-\\frac{5\\sqrt{5}}{8\\sqrt{3}}$e^{{-2h}}$+O($e^{{-4h}}$).$$ Since the curve traced by the unstable eigenvalues is approximated by an ellipse of horizontal width proportional to $|\\beta_1^{(p)}|\\epsilon^p$, these formulas give the exact exponential rates at which the $p=2,3,4$ isolas shrink as the depth grows. The nonzero constants imply that each of these three coefficients is nonzero for all sufficiently large $h$, and, being real analytic, each has at most finitely many positive-depth zeros; the proof uses the explicit 27-term (for $p=4$) algebraic expressions for $\\beta_1^{(p)}$ and refined asymptotics for the branching wavenumber.","pith_inferences":["Beyond the paper, the same algebraic formulas could be expanded at finite depth to produce rigorous two-sided bounds on the isola widths, turning the numerical spectral plots into certified existence statements for each $p$.","The decay rates imply a concrete challenge for numerical solvers: for a tank of depth $h$, the $p=2,3,4$ eigenvalue gaps are of order $e^{-h/2}$ to $e^{-2h}$ times $\\epsilon^p$, so direct detection requires resolving exponentially small separations.","The clean form of the $p=4$ constant, a rational multiple of $\\sqrt{15}$, hints that the unshown cancellation in (2.56) may admit a human-readable derivation, which would make the $p=4$ case fully checkable without computer algebra."],"forward_implications":["For $p=2,3,4$, $\\beta_1^{(p)}(h)$ is nonzero at all sufficiently large depths, so the corresponding high-frequency instability isolas exist for all deep enough water and do not accumulate an infinite sequence of critical depths.","The Floquet interval supporting the p-th isola has width $4|\\beta_1^{(p)}|/T_1^{(p)}\\,\\epsilon^p + O(\\epsilon^{p+1})$, hence in deep water it shrinks like $e^{-h/2}$ for $p=2$ and like $e^{-2h}$ for $p=3,4$.","The maximum growth rate of the unstable mode, of size $|\\beta_1^{(p)}|\\epsilon^p$, is exponentially small in the depth for these three bands, making the deep-water instability a weak but persistent effect.","Since $h\\mapsto\\beta_1^{(p)}(h)$ is real analytic, the expansions prove Conjecture 2.2 for $p=2,3,4$: each of these coefficients has only finitely many zeros in $(0,\\infty)$."],"supporting_citations":[{"why":"Supplies Theorem 1.1, the explicit formulas (5.7)–(5.8) for $\\beta_1^{(p)}$, and the prior limits $\\beta_1^{(p)}\\to-\\infty$ as $h\\to0^+$ and $\\beta_1^{(p)}\\to0$ as $h\\to+\\infty$.","marker":"[3]"},{"why":"Gives the formulas (2.20)–(2.23) for the Stokes-wave coefficients $p_1^{[1]}$, $p_2^{[2]}$, $a_1^{[1]}$, $a_2^{[2]}$ used in the $p=2,3$ expansions.","marker":"[5]"},{"why":"Gives (A.59)–(A.60) for the third- and fourth-order coefficients $p_3^{[3]}$, $a_3^{[3]}$, $p_4^{[4]}$, $a_4^{[4]}$ used for $p=3,4$.","marker":"[6]"},{"why":"Proved the existence of the $p=2$ isola in deep water, the case whose asymptotic expansion (2.1) sharpens.","marker":"[7]"},{"why":"Developed the formal perturbation expansion of the first two high-frequency isolas that the $p=2,3$ computations here justify in the deep-water limit.","marker":"[13]"},{"why":"Provided the numerical discovery and conjecture of infinitely many shrinking isolas that the theorem confirms quantitatively for $p=2,3,4$.","marker":"[16]"}],"fun_headline_variants":["Exact decay rates for deep-water Stokes isolas","Explicit exponential decay of p=2,3,4 Stokes isolas","Deep-water asymptotics pin down Stokes instability isolas","Stokes isolas shrink exponentially in deep water","New rates for deep-water decay of Stokes instability bands"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The $p=4$ expansion rests on a long computer-algebra cancellation, not written out in the paper, together with the correctness of the 27-term formula taken from the companion paper; if either is wrong, the stated leading constant $-\\frac{5\\sqrt{5}}{8\\sqrt{3}}$ in (2.3) would change.","fun_headline_variants_meta":{"raw":{"variants":["Exact decay rates for deep-water Stokes isolas","Explicit exponential decay of p=2,3,4 Stokes isolas","Deep-water asymptotics pin down Stokes instability isolas","Stokes isolas shrink exponentially in deep water","New rates for deep-water decay of Stokes instability bands"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000577,"raw_usage":{"total_tokens":2770,"prompt_tokens":1041,"completion_tokens":1729,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":657,"completion_tokens_details":{"reasoning_tokens":1649}},"tokens_in":657,"tokens_out":1729,"duration_ms":12354,"temperature":1.0,"reasoning_tokens":1649,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:29:01.029765+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the signed sum in (2.56) with independent symbolic or high-precision arithmetic: if the eight three-intermediate-harmonic terms do not cancel to $O(e^{-4h})$, the $p=4$ statement falls. Alternatively, evaluate $\\beta_1^{(4)}(h)$ numerically at large $h$ (for example $h=20$) using the formulas of Section 2.3 and compare with $-\\frac{5\\sqrt{5}}{8\\sqrt{3}}e^{-2h}$; a mismatch beyond the stated $O(e^{-4h})$ error would refute Theorem 2.1.","supporting_citations":[{"cited_title":"Berti, A","cited_arxiv_id":null,"evidence_quote":"Gives the formulas (2.20)–(2.23) for the Stokes-wave coefficients $p_1^{[1]}$, $p_2^{[2]}$, $a_1^{[1]}$, $a_2^{[2]}$ used in the $p=2,3$ expansions."},{"cited_title":"Berti, A","cited_arxiv_id":null,"evidence_quote":"Gives (A.59)–(A.60) for the third- and fourth-order coefficients $p_3^{[3]}$, $a_3^{[3]}$, $p_4^{[4]}$, $a_4^{[4]}$ used for $p=3,4$."},{"cited_title":"First isola of modulational instability of Stokes waves in deep water","cited_arxiv_id":"2401.14689","evidence_quote":"Proved the existence of the $p=2$ isola in deep water, the case whose asymptotic expansion (2.1) sharpens."},{"cited_title":"Creedon, B","cited_arxiv_id":null,"evidence_quote":"Developed the formal perturbation expansion of the first two high-frequency isolas that the $p=2,3$ computations here justify in the deep-water limit."},{"cited_title":"Deconinck and K","cited_arxiv_id":null,"evidence_quote":"Provided the numerical discovery and conjecture of infinitely many shrinking isolas that the theorem confirms quantitatively for $p=2,3,4$."}],"review_version":1}