{"id":"4784c201-8c4c-4fdf-b05c-b0959beeec8e","arxiv_id":"2603.12134","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Finite element magnetic relaxation that preserves only global helicity cannot maintain braided topology that all-local-helicity preservation keeps, and this global-only scheme may emulate Taylor relaxation.","lead":"Magnetic relaxation simulations often need to conserve helicity to avoid unphysical outcomes. This paper shows that conserving only total helicity, via a Lagrange multiplier, still lets numerical reconnections destroy braided magnetic topology, whereas conserving helicity in every subdomain preserves it—and the global-only scheme may actually mimic resistive Taylor relaxation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Local-helicity conservation (Theorem 4.2) is the load-bearing step; its proof uses an inadmissible H0(curl) test function and an undefined discrete potential, though the gap is likely repairable.","rationale":"The reader's conditional verdict is well-aimed: Theorem 4.2 is the linchpin of the paper's central claim, and the proof as written has a genuine gap. I agree that using A_h as a test function in the H0(div) Faraday equation is not justified by the stated function spaces. However, I do not think the gap is necessarily fatal. The residual in (4.2a) lies in H_h(div), so the strong Faraday law likely holds and one can multiply by A_h a posteriori. The larger missing piece is a discrete potential family: the proof assumes identities relating A_h^{n+1}, A_h^n, B_h^{n+1}, and B_h^n that require A_h^n to evolve consistently with the Faraday update, but Problem 4.1 never introduces A_h^n as a variable. This is a fixable omission rather than a demonstrated contradiction. The numerical code is archived, so the conservation law can be checked directly; that check would settle whether the theorem's statement is true. I did not find a separate concern strong enough to change the verdict: the Lagrange-multiplier scheme and the numerical comparisons are still worth conditional acceptance, with the burden on the authors to either supply the missing discrete-potential argument or verify (4.4) computationally.","tokens_in":16860,"tokens_out":12530,"duration_ms":119743,"concrete_test":"Use the archived Firedrake code [21] to run one time step of Problem 4.1 on the 4x4x10 Hopf mesh. Construct A_h^n by solving the discrete curl-inverse problem curl A_h^n = B_h^n in the Nédélec space with a gauge condition, and similarly for A_h^{n+1}. For every union of cells Ω_s,h whose B_h boundary normal trace is zero, compute H_h^n(Ω_s,h)=∫_{Ω_s,h} A_h^n·B_h^n and H_h^{n+1}(Ω_s,h). Check whether (4.4) holds to machine precision. If it fails, Theorem 4.2 is false as stated and the central claim collapses; if it holds, the proof gap is expository and the reader's concern is not fatal.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central distinction—projection-based local helicity preservation vs. Lagrange-multiplier global-only preservation—rests on Theorem 4.2, specifically equation (4.4). As written, the proof takes C_h = A_h^{n+1/2} in (4.2a), but C_h is required to lie in H0^h(div), while A_h is a Nédélec function in H0^h(curl). For lowest-order edge and face elements these spaces are not nested: Nédélec functions need not have continuous normal traces, so the test function is inadmissible. The gap is not merely cosmetic. Even if one recovers the strong form of Faraday's law (the residual lies in H_h(div), so (4.2a) implies ∂tB_h = -curl E_h), the displayed algebra also needs a time-consistent family of discrete potentials A_h^n satisfying A_h^{n+1} = A_h^n - Δt E_h^{n+1/2} + ∇s_h, so that ∫(A^{n+1}-A^n)·B^{n+1/2} = ∫(B^{n+1}-B^n)·A^{n+1/2}. Problem 4.1 does not define A_h^n or its evolution, and no argument is given for why the restriction has zero tangential trace on each discrete magnetically closed subdomain. Since the interpretation of Figures 6.2 and 6.4 depends on (4.4), a proof gap here is load-bearing. The statement may be repairable, but the manuscript does not supply the repair.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript investigates how the level of discrete magnetic-helicity conservation influences the relaxed states computed for the magneto-frictional (MF) equations. Three finite element schemes are compared: a non-conservative discretization, the authors' earlier projection-based scheme that is claimed to preserve helicity on every discrete magnetically closed subdomain, and a newly introduced Lagrange multiplier scheme that enforces only global helicity. The paper proves energy decay, the discrete Gauss law, and a discrete Arnold inequality for the new scheme, and claims local helicity conservation for the projection scheme in Theorem 4.2. Numerical experiments on an E3 braid (zero helicity) and a Hopf knot (nonzero helicity) are used to argue that global-only helicity conservation leads to qualitatively different—and for the braid, trivial—relaxed states, and that the Lagrange multiplier scheme may serve as a numerical model of Taylor relaxation.","tokens_in":17356,"tokens_out":14675,"duration_ms":125781,"significance":"The question addressed—whether global helicity conservation alone is sufficient for faithful numerical magnetic relaxation, or whether local helicity constraints are essential—is timely and consequential. If the results are correct, the paper would demonstrate a striking dependence of the relaxed state on the level of discrete topological constraint and would offer a computationally cheap scheme for simulating Taylor-like relaxation. The manuscript is honest about its limitations: Remark 6.2 explicitly states that the nontrivial steady state for the E3-field is not proven, and Section 7 disclaims any definitive answer on the Taylor-relaxation interpretation. The numerical code is archived on Zenodo. However, the central theoretical result on local helicity conservation is not proven as written, and the E3 conclusion rests on a single coarse-mesh experiment.","major_comments":[{"comment":"The proof of local helicity conservation takes C_h = A_h^{n+1/2} in the Faraday equation (4.2a), but C_h is required to lie in H^h_0(div), whereas A_h^{n+1/2} is only in H^h_0(curl). In the FEEC de Rham complex these are different spaces (Nédélec edge vs. Raviart–Thomas face) with no inclusion, so the test function is inadmissible as written. Moreover, the potentials A_h^n appearing in (4.4) are not defined in Problem 4.1, and no argument is given for the required tangential-trace conditions on ∂Ω_{s,h}. Since (4.4) is the basis for distinguishing the projection-based scheme from the Lagrange multiplier scheme, this gap is load-bearing. A repair should either derive the strong form of Faraday's law (which does follow from (4.2a) because the test space spans H^h_0(div)) and then integrate by parts, or construct time-consistent potentials via a discrete Poincaré operator with a controlled","section":"Section 4, Theorem 4.2, Eq. (4.2a), (4.4)"},{"comment":"The conclusion that the Lagrange multiplier scheme converges to the trivial state for the E3-field is supported by a single coarse mesh (4×4×24). Since any consistent discretization of the ideal MF equations should approach the local-helicity-preserving solution as h → 0, the observed decay may reflect resolution-dependent numerical diffusion rather than a fundamental property of global-helicity enforcement. A mesh-refinement study, or at least a quantitative measure of local-helicity violation for the Lagrange multiplier scheme, is needed before drawing the qualitative conclusion stated in Section 1. This is especially important because Remark 6.2 already concedes that the nontrivial steady state of the projection scheme is not rigorously established.","section":"Section 6.1, Figures 6.1–6.2"},{"comment":"The symbol E^{n+1}_h is used both for the electric field unknown and for the magnetic energy in the discrete energy law (5.10f). The unknown list at the top of Problem 5.3 includes E^{n+1}_h as an H(curl) field, but (5.10f) is a scalar equation involving the energy. This overloading makes the formulation ambiguous and should be corrected (e.g., use ℰ_h for the magnetic energy). The same ambiguity appears in the continuous derivation in Section 5.1.","section":"Section 5.2, Problem 5.3 and Eq. (5.10f)"}],"minor_comments":[{"comment":"The displayed statement 'H^{n+1}_h = (A^n_h, B^n_h)' appears to be a typo; it should likely be 'H^n_h = (A^n_h, B^n_h)'.","section":"Theorem 4.2"},{"comment":"The gauge invariance of local helicity ∫_Ωs A·B on a magnetically closed subdomain is asserted but not stated precisely. A short explanation of the boundary-term cancellation would help the reader.","section":"Section 2, Eq. (1.2)"},{"comment":"The 'Error' panels plot quantities labeled H_h and ‖∇·B_h‖, but the captions do not specify whether these are absolute errors, relative errors, or solver tolerances. Please clarify.","section":"Figure captions 6.1 and 6.3"},{"comment":"The switching threshold γ = 9×10^{-5} in the Lagrange multiplier scheme is a free parameter; no sensitivity study is provided. A brief remark on its influence on the reported steady states would strengthen the numerical section.","section":"Remark 5.5"}],"recommendation":"major_revision","confidential_remarks":"The gap in Theorem 4.2 is serious but likely repairable; the authors should be asked to provide a correct proof or to clearly mark the local-helicity conservation as conditional. The E3 numerical claim needs a resolution study or a softened wording. The paper is otherwise well within the journal's scope and the new Lagrange multiplier scheme is a useful contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe paper you'll care about: it compares three FE schemes for magneto-frictional relaxation — no constraint, projection-based local helicity preservation, and a new Lagrange multiplier enforcing only global helicity — on the E3 braid and the Hopf knot. The new Lagrange multiplier scheme is the main contribution; the comparison showing that global-only conservation can relax to a different (or trivial) state than local conservation is genuinely informative. The authors are also honest: Remark 6.2 explicitly says the E3 nontrivial-state claim is not proven, and Section 7 frames the Taylor-relaxation interpretation as an open question. Code is archived on Zenodo, which earns credit.\n\nThe soft spots are real but probably repairable. Theorem 4.2 is the load-bearing step: local helicity conservation in the projection scheme. As written, the proof chooses C_h = A_h^{n+1/2} as a test function in the Faraday equation (4.2a), but C_h lives in H0^h(div) while A_h is a Nédélec function in H0^h(curl). For lowest-order edge and face elements these spaces don't nest, so the test function is inadmissible. The proof also assumes a time-consistent discrete potential A_h^n satisfying A^{n+1}=A^n - Δt E^{n+1/2}, but Problem 4.1 never defines A_h^n and no such evolution is established. The conservation statement may well be true — it likely follows from the structure of the scheme, and a similar argument exists in [29] — but as written the proof is incomplete. Since the central distinction between the projection and Lagrange multiplier schemes depends on (4.4), this gap should be patched before publication.\n\nThe numerical experiments are suggestive but not conclusive: everything runs on a single coarse mesh (4×4×10 or 4×4×24), and the Lagrange multiplier scheme has a hand-set switching threshold γ. No mesh convergence study is included, so the qualitative differences in equilibria need confirmation. That's a moderate concern, not a fatal one.\n\nOverall: this is a competent, honest paper that should see the inside of a referee's office. I'd recommend conditional acceptance with a required repair of Theorem 4.2's proof and, ideally, at least one mesh-refinement check. The speculative Taylor-relaxation framing is clearly labeled and can stay as discussion.\n\nBest.","headline":"A useful numerical comparison of global vs. local helicity constraints in magnetic relaxation, with an interesting new Lagrange multiplier scheme, but the proof of local helicity conservation is incomplete and the experimental claims rest on one coarse mesh.","tokens_in":17715,"tokens_out":2922,"would_cite":true,"duration_ms":26270,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65N30","65L60","76W05"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that local helicity conservation in finite element magnetic relaxation is necessary: enforcing only the global helicity lets spurious reconnection drive the field to the wrong steady state.","keywords":["magnetic relaxation","magnetic helicity","local helicity conservation","finite element exterior calculus","structure-preserving methods","magneto-frictional equations","magnetic braids","global helicity constraint"],"falsifier":"Run the projection-based scheme on the E3 braid and measure the computed local helicity on a single hexahedral cell (or any discrete magnetically closed subdomain) over thousands of time steps; a drift larger than the solver tolerance would falsify the claimed local conservation. Conversely, a fine-mesh computation with the Lagrange multiplier scheme that relaxes the E3 braid to a nontrivial steady state would falsify the claim that global-only conservation always yields trivial relaxation.","tokens_in":16788,"feed_emoji":"🧲","tokens_out":8021,"duration_ms":60477,"temperature":0.7,"pith_summary":"Magnetic relaxation, the process by which a plasma lowers its magnetic energy while preserving topological constraints, is a prime testing ground for structure-preserving finite element methods. This paper compares three discretizations of the magneto-frictional equations: an unconstrained scheme, a projection-based scheme that conserves helicity on every magnetically closed subdomain, and a new Lagrange multiplier scheme that enforces only the total helicity. The central finding is that the level of helicity preservation changes the relaxed state: the unconstrained scheme drives the field to zero, the global-only scheme destroys braided structure and reaches a different nonzero equilibrium for knots, while the local scheme preserves nontrivial topology and reaches a stable nontrivial equilibrium even for a zero-helicity braid. The paper argues that the global-only scheme, despite its numerical reconnections, may serve as a model of relaxation theories in which only total helicity survives.","feed_headline":"Local helicity sets the relaxed magnetic state","feed_subtitle":"A finite element comparison shows that conserving helicity on every subdomain, not just globally, decides the equilibrium.","key_machinery":"The central object is the discrete magnetic helicity on magnetically closed subdomains, together with the discrete helicity-energy bound that prevents the energy from collapsing to zero when helicity is nonzero. Two constructions carry the argument. The first is the projection scheme, which projects the face-element magnetic field into an edge-element curl space through an auxiliary variable; this makes the local helicity a discrete invariant and yields the discrete lower bound. The second is the Lagrange multiplier scheme, which enforces only the global helicity via two real scalar multipliers, allowing local reconnection while keeping the total helicity constant. The contrast between these","core_discovery":"The paper's central claim is that discrete conservation of local helicity on every magnetically closed subdomain is what prevents spurious numerical reconnection and yields the physically correct relaxed state of the ideal magneto-frictional system. This is established by comparing three schemes. The projection-based scheme introduces an auxiliary magnetic potential in the curl space, which makes the local helicity on any discrete magnetically closed subdomain a conserved invariant and satisfies a discrete helicity-energy bound. The Lagrange multiplier scheme enforces only the global helicity as a scalar constraint, allowing local topology to change through numerical reconnection. Numerical","pith_inferences":["A direct test of conservation on a single finite element or small patch would either verify or falsify the paper's central distinction; the proof as written leaves this gap because the discrete vector potential belongs to a different finite element space than the test functions in the Faraday equation.","If the local-conservation proof is repaired, the same auxiliary-variable technique could be applied to other finite element exterior calculus discretizations, such as extended MHD models, to obtain local helicity preservation there.","The paper leaves open whether configurations with zero local but higher-order linking (for example, Borromean rings) would require constraints beyond local helicity; a local-helicity-conserving scheme may still permit changes in triple linking.","For solar coronal heating studies, the mesh size or time step of the global-only scheme could be tuned to control the rate of numerical reconnection, turning discretization error into a proxy for physical resistivity."],"forward_implications":["If the local conservation result holds, structure-preserving simulation of ideal magnetic relaxation must conserve helicity on every magnetically closed subdomain, not just the domain as a whole.","The failure of the global-only scheme on the zero-helicity braid shows that total helicity is not a sufficient topological barrier when the initial field has zero net helicity.","The distinct nonzero equilibria for the Hopf knot imply that the computed relaxed state depends on which invariants are preserved, so numerical relaxation studies should state their conservation properties explicitly.","The Lagrange multiplier scheme, though less faithful to the ideal equations, qualitatively reproduces the behavior assumed in relaxation theories with only a global constraint, suggesting a deliberate use for resistive relaxation.","The discrete helicity-energy bound holds for both structure-preserving schemes, guaranteeing a minimum energy whenever the initial helicity is nonzero."],"fun_headline_variants":["Local helicity, not global, decides relaxed plasma state","Preserving local helicity stops spurious reconnection","Magnetic relaxation: local helicity locks the topology","Helicity per subdomain controls magnetic relaxation outcomes","Local conservation beats global in helicity-preserving FEM"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof of local helicity conservation assumes that the discrete vector potential, which lives in a curl-conforming edge-element space, can be used as a test function in the div-conforming Faraday equation; because these two spaces do not nest, this step is not justified as written, and if local conservation fails on some discrete magnetically closed subdomain, the central distinction between the two schemes collapses.","fun_headline_variants_meta":{"raw":{"variants":["Local helicity, not global, decides relaxed plasma state","Preserving local helicity stops spurious reconnection","Magnetic relaxation: local helicity locks the topology","Helicity per subdomain controls magnetic relaxation outcomes","Local conservation beats global in helicity-preserving FEM"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000159,"raw_usage":{"total_tokens":1021,"prompt_tokens":653,"completion_tokens":368,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":397,"completion_tokens_details":{"reasoning_tokens":292}},"tokens_in":397,"tokens_out":368,"duration_ms":3837,"temperature":1.0,"reasoning_tokens":292,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T18:18:57.313853+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the projection-based scheme on the E3 braid and measure the computed local helicity on a single hexahedral cell (or any discrete magnetically closed subdomain) over thousands of time steps; a drift larger than the solver tolerance would falsify the claimed local conservation. Conversely, a fine-mesh computation with the Lagrange multiplier scheme that relaxes the E3 braid to a nontrivial steady state would falsify the claim that global-only conservation always yields trivial relaxation.","supporting_citations":[],"review_version":1}