{"id":"d48aecfb-071a-47b2-84a0-3a69094f9f67","arxiv_id":"1908.07148","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In 2D MHD simulations of solar filament longitudinal oscillations with gravity comparable to the Lorentz force, wave leakage and non-adiabatic losses shorten the decay time and make the pendulum model's curvature estimate about 100 percent off.","lead":"A simulation of solar filament oscillations in weak magnetic fields finds that the oscillating plasma deforms the field and radiates waves that drain its kinetic energy. The result suggests the common pendulum-based method for measuring magnetic dip curvature from observed periods can be roughly 100 percent off in this regime.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The ~100% pendulum-model error claim conflates thermal period shifts with field deformation: the paper compares 1D non-adiabatic P=30 min to 2D non-adiabatic P=49 min, while cooling alone shortens the 1D period from 37 to 30 min.","rationale":"The reader's weakest assumption was the 2D-to-3D extrapolation, which is a genuine external-validity concern. However, the more immediate internal problem is that the quantitative headline rests on a period comparison that mixes thermal effects with field deformation. The paper reports four periods: 1D adiabatic 37 min, 1D non-adiabatic 30 min, 2D adiabatic 44 min, and 2D non-adiabatic 49 min. Section 4.3 uses the 30 and 49 values to infer the 100% curvature error. Since the 1D non-adiabatic case is not a pendulum model, because it includes cooling-induced pressure changes, comparing it to the 2D non-adiabatic case overstates the deformation effect. The adiabatic pair is the cleaner control and gives a much smaller effect. This is a concrete, falsifiable flaw in the central quantitative claim, independent of whether the 3D piston argument is correct. The simulation study remains useful and the wave-leakage mechanism plausible, so the conditionally-accept verdict stands, but the paper should either weaken the 100% claim or add the missing baseline runs.","tokens_in":13478,"tokens_out":17106,"duration_ms":188860,"concrete_test":"Re-run the 1D simulation with the initial field line but without radiation and heat conduction (the adiabatic case already gives P=37 min), and compare that period and the pure-gravity pendulum period computed from the local curvature radius of the initial centroid field line to the 2D adiabatic period (44 min) and the 2D non-adiabatic period (49 min). If the deformation-induced period ratio is about 44/37, the curvature-radius error is about 40%, not 100%. Report fits over the same first-cycle interval with bootstrap uncertainties.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4.3 derives the headline claim by comparing the 2D non-adiabatic period (49 min) with the 1D non-adiabatic period (30 min) and attributing the 19-minute increase to magnetic-field deformation, leading to a >50% period error and ~100% curvature-radius error. This comparison does not isolate deformation. In the same paper, the 1D adiabatic period is 37 min, so switching on radiation and heat conduction shortens the rigid-field period by 7 min (19%); the 2D adiabatic period is 44 min, only 7 min longer than the 1D adiabatic period. Thus the 30-to-49 difference used for the 100% error includes roughly as much thermal-induced period shortening as deformation-induced lengthening. A clean deformation test would compare 2D adiabatic (44) with 1D adiabatic (37), giving a period increase of about 19% and a curvature error of about 40%, not 100%. Additionally, both signals are chirps, as the paper itself notes that the period decreases with time, so the fitted constant periods depend on the fitting window and are quoted without uncertainties. The abstract's '~100%' statement is therefore not established by the simulations as presented; a proper pendulum baseline (pure gravity on the initial field line, or the adiabatic rigid-field period) is needed before this number is used for seismology.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"Zhang, Fang, and Chen present two-dimensional (2D) MHD simulations of a filament thread oscillating longitudinally in a dipped quadrupolar magnetic field with gravity-to-Lorentz force ratio δ close to unity. They compare 2D non-adiabatic, 2D adiabatic, 1D non-adiabatic, and 1D adiabatic runs, fitting the field-aligned velocity with a damped sine to extract period and decay time. The central findings are that non-adiabatic processes (radiation and heat conduction) reduce the decay time, that the 2D runs lose additional energy through outgoing fast-mode wave trains generated by magnetic-field deformation, and that in this regime the pendulum model may misestimate the curvature radius of the magnetic dip by about 100%.","tokens_in":13929,"tokens_out":6108,"duration_ms":60970,"significance":"If established, the paper would make two useful contributions: a quantitative demonstration that wave leakage can shorten filament longitudinal oscillation decay times when the magnetic field is weak, and a warning that the pendulum model used for prominence seismology may fail when gravity and Lorentz force are comparable. The strength of the paper is its concrete evidence for the wave-leakage channel: the time-distance diagram in Figure 11 shows quasi-periodic vertical-velocity disturbances propagating upward and downward at the coronal fast-mode speed, and the energy-budget decomposition in Figure 12 indicates that Lorentz-force and pressure work remove a major part of the initial kinetic energy. The δ parameter adopted from Zhou et al. (2018) gives a useful ordering scheme for when field deformation matters. However, the headline 100% curvature-radius error is not currently established, because it is based on a comparison that mixes thermal period shifts with magnetic deformation, and the paper contains several internal numerical inconsistencies in the reported periods and decay times.","major_comments":[{"comment":"The claim that the pendulum model leads to an error of ~100% in the curvature radius is not established by the comparison made in the text. The paper compares the 2D non-adiabatic period (P = 49 min) with the 1D non-adiabatic period (P = 30 min) and attributes the 19-minute difference to magnetic-field deformation. But the paper's own numbers show that switching on radiation and heat conduction shortens the 1D rigid-field period from 37 min (adiabatic) to 30 min, a 19% thermal shift, while the 2D adiabatic period is 44 min, only 19% longer than the 1D adiabatic period. A deformation-only comparison (2D adiabatic 44 min vs. 1D adiabatic 37 min) gives a period increase of about 19% and a curvature-radius error of about 40%, not 100%. In addition, the signals are chirped (the paper itself notes the period decreases with time), so the constant fitted periods depend on the fitting window and are quoted without uncertainties. A proper rigid-field adiabatic baseline, or a time-resolved period estimate, is needed before the abstract's '~100%' statement is used.","section":"§4.3 and Abstract"},{"comment":"The decay times reported for the same cases are internally inconsistent. Section 3.2 gives τ = 76 min for the 1D non-adiabatic case, but Summary item (3) states the decay time is reduced from 113 min in 1D to 34 min in 2D. Section 3.2 gives τ = 38 min for the 2D non-adiabatic case, while the text in Section 4 and Summary items (2) and (3) use 34 min. Section 3.3 gives τ = 211 min for the 2D adiabatic case, but Section 4.2 states that the 2D adiabatic case has a decay time of 76 min. These are exactly the numbers used to quantify the factor-of-two reduction and the τ/P = 0.7 claim, so they must be reconciled and the quoted values corrected consistently throughout.","section":"§3.2, §3.3, §4, and Summary"},{"comment":"The extrapolation from the 2D slab result to real 3D filaments is asserted rather than demonstrated. The paper argues that the filament's sheared core field overlain by an unsheared envelope means the oscillating filament acts as a piston even in 3D, so wave leakage remains efficient. However, no 3D simulation or quantitative estimate of the piston efficiency is provided, and the paper itself notes that Terradas et al. (2006) found wave leakage ineffective for 3D flux tubes in an aligned ambient field. If, in 3D, the ambient field lines are pushed aside rather than forced to oscillate, the wave-leakage channel and the associated seismology error could be substantially smaller. Because the abstract's general claim about real filaments depends on this extrapolation, the paper should either add a 3D test, provide a quantitative geometric estimate, or explicitly restrict the claim to the 2D slab configuration.","section":"§4.2, final paragraph"}],"minor_comments":[{"comment":"There is a typo: 'tha application of the pendulum model' should read 'the application of the pendulum model.'","section":"Summary, item (1)"},{"comment":"The text refers to 'shifting Tp' in the decomposition of the pressure-gradient force, but the variable is Fp throughout the section; this should be corrected.","section":"§4.1, Figure 7 discussion"},{"comment":"The caption contains 'Time-distance diagram of of the temperature distribution'; the duplicate 'of' should be removed.","section":"Figure 4 caption"},{"comment":"The damped-sine fits are described as reasonable for only the first 1.5 periods, with deviations becoming remarkable afterward, yet the extracted periods and decay times are quoted without uncertainties or a goodness-of-fit measure; reporting the fitting window and uncertainties would make the quantitative comparisons more robust.","section":"§3.1 and §3.2"},{"comment":"The two side boundaries are reflecting; a brief discussion of whether waves reflected from these boundaries affect the measured decay times would help the reader assess the numerical setup.","section":"§2, boundary conditions"}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the journal's scope and the wave-leakage mechanism is a genuinely interesting and potentially important result for prominence seismology. My recommendation of major revision is driven by two issues: the headline 100% pendulum-model error is obtained from a comparison that does not isolate magnetic deformation, and the paper has multiple inconsistent values for the decay times of the same simulations. These are fixable within the manuscript's scope by redoing the comparison with a proper rigid-field baseline and correcting the numbers. I do not see a reason to question the authors' good faith, and the direct wave-train evidence in Figure 11 is a strong positive element."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real simulation study with a plausible new result, but the headline \"~100% pendulum error\" does not survive contact with the paper's own numbers. The 2D non-adiabatic run in the delta~1 regime shows outgoing fast-mode wave trains and a much shorter decay time than the 1D runs, and Figure 12's energy decomposition supports wave leakage as the extra channel. That is genuinely new and worth building on.\n\nThe paper's design is sensible: four cases (1D/2D x adiabatic/non-adiabatic) with the same initial field-aligned perturbation, and the pressure-gradient analysis in Section 4.1 gives a clean physical picture of why radiation and conduction turn the pressure force into a viscous drag. The fast-mode wave signature in Figure 11 (period 6.38 min, speed 810 km/s) is direct evidence. I believe the central mechanism.\n\nThe abstract's ~100% claim, however, is not established. Section 4.3 attributes the 19-minute period increase from 1D non-adiabatic (30 min) to 2D non-adiabatic (49 min) to field deformation, but the same paper shows the 1D adiabatic period is 37 min and the 2D adiabatic is 44 min. So thermal effects alone shorten the rigid-field period by 7 minutes; comparing 2D adiabatic with 1D adiabatic isolates deformation and gives about a 19% period increase, i.e., roughly 40% curvature-radius error, not 100%. The stress-test note is right. Both signals are chirps, and no uncertainties are given for the fitted P and tau, so the quoted numbers are point estimates from an underspecified fit.\n\nThere are also internal inconsistencies in the damping times that need reconciliation: the 1D non-adiabatic tau is 76 min in Section 3.2, but 113 min in the Section 5 summary; the 2D non-adiabatic tau is 38 min in Section 3.1 and 34 min in the summary. That is the kind of thing a careful referee will trip on.\n\nThe 2D-to-3D extrapolation in Section 4.2 is honest—they cite Terradas et al. and argue the filament's sheared-core/unsheared-envelope geometry makes a piston even in 3D—but it is an argument, not a simulation, so the real-world seismology impact remains open.\n\nBottom line: this paper deserves peer review and probably publication after the period-error comparison is redone with a proper baseline and the decay-time numbers are reconciled. I would not trust the ~100% number as it stands, but the wave-leakage mechanism for delta~1 is solid enough to cite.","headline":"Solid wave-leakage simulation undermined by an overstated ~100% claim that mixes thermal and deformation effects.","tokens_in":14362,"tokens_out":3763,"would_cite":true,"duration_ms":32580,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In weak magnetic fields, oscillating solar filaments leak fast-mode waves, halve their decay time, and break the pendulum-model relation between period and dip curvature — overestimating the inferred curvature radius by ~100%.","keywords":["magnetohydrodynamics (MHD)","solar filaments","prominences","longitudinal oscillations","wave leakage","prominence seismology","pendulum model","fast-mode waves"],"falsifier":"A 3D MHD simulation of a filament with $\\delta\\approx 1$ and a realistic sheared-core/unsheared-envelope field would settle it: if the decay time is not roughly half the matched 1D value and no fast-mode wave trains stream away from the filament, the wave-leakage claim fails. Observational support could come from detecting quasi-periodic 810 km/s wavefronts propagating above oscillating filaments in weak-field regions; their absence in many events would indicate the 2D piston picture does not apply.","tokens_in":13301,"feed_emoji":"🌞","tokens_out":13335,"duration_ms":113587,"temperature":0.7,"pith_summary":"The paper tries to establish that two-dimensional and non-adiabatic effects, not just gravity along a rigid field line, control how solar filament longitudinal oscillations damp. Using 2D MHD simulations in a regime where gravity is comparable to the Lorentz force ($\\delta\\approx 1$), it shows that the oscillating filament deforms its supporting magnetic field by about 0.5 Mm, launches fast-mode wave trains away from the filament, and damps roughly twice as fast as 1D simulations predict. It further claims that in this regime the standard pendulum relation between period and dip curvature radius is not reliable, overestimating the curvature radius by about 100%. The paper matters because filament oscillations are used as a seismological tool to infer magnetic field geometry, and because observed decay times are often shorter than 1D models can explain.","feed_headline":"Leaking waves double the seismology error for weak-field filaments","feed_subtitle":"2D MHD simulations show oscillating filaments shed fast-mode waves, halving decay times and breaking the pendulum model.","key_machinery":"The load-bearing object is the dimensionless gravity-to-Lorentz ratio $\\delta = \\rho g L/(B^2/2\\mu_0)$; when $\\delta$ is near unity the dense filament can deform the magnetic field rather than sliding along a rigid tube. A second element is the pendulum relation $P=2\\pi\\sqrt{R/g}$, which the paper tests and finds wanting in that regime. The mechanism that carries the 2D damping is the piston effect: in a 2D slab the oscillating filament pushes all nearby field lines, generating transverse oscillations and outgoing fast-mode magnetoacoustic waves. The paper also decomposes the gas-pressure gradient force into a part antiphase with velocity (a viscous damping force in non-adiabatic runs) and a residual restoring part, explaining why radiation and heat conduction convert pressure forces from restoring to damping.","core_discovery":"On its own terms, the paper reports that in a 2D non-adiabatic MHD simulation with the gravity-to-Lorentz ratio $\\delta$ close to unity, a perturbed filament thread oscillates with period $P\\approx 49$ minutes and decay time $\\tau\\approx 38$ minutes ($\\tau \\approx 0.7P$), whereas the matched 1D non-adiabatic run gives $P\\approx 30$ minutes and $\\tau\\approx 76$ minutes ($\\tau \\approx 2.5P$). The 2D adiabatic run decays in about 211 minutes while the 1D adiabatic run is essentially decayless, isolating wave leakage as the extra 2D damping channel. The deformation of the field line by the moving filament excites a transverse oscillation with period $\\approx 6.38$ minutes and quasi-periodic fast-mode waves that carry energy upward and downward at about 810 km/s; energy-budget integrals attribute the loss to Lorentz force and gas pressure work, i.e., to wave radiation. Because the field flattens as the filament moves and the period lengthens by over 50% relative to 1D, the pendulum model's inferred dip curvature radius is wrong by roughly 100% in this regime.","pith_inferences":["If the 2D piston picture survives in 3D for the sheared-core/unsheared-envelope geometry, then filament oscillations in weak-field regions should be observable sources of outward-propagating quasi-periodic coronal disturbances; a systematic search near oscillating filaments would test this.","The period drift predicted here (the period shortens as the oscillation decays) could serve as an observational diagnostic of the $\\delta$ regime even when the magnetic field strength is not directly measurable.","The roughly 100% error bound is likely an upper limit for real filaments: in 3D geometries where ambient field lines slip sideways around the flux tube, wave leakage weakens and the pendulum model partially recovers.","Energy carried away by leaked fast-mode waves is deposited in the ambient corona, so filament oscillations may contribute to local coronal heating near weak-field filaments, a consequence the paper does not quantify."],"forward_implications":["Observed decay times shorter than 1D predictions can be explained by wave leakage plus radiation and conduction, without needing mass drainage or thread-thread interaction.","In weak-field events ($\\delta\\approx 1$), measured periods should not be fed directly into the pendulum formula; inferred dip curvature radii will be roughly twice too large.","Longitudinal oscillations in this regime should be accompanied by small-amplitude transverse oscillations with periods of minutes and by upward and downward fast-mode wave trains in the surrounding corona.","The predicted $\\tau/P \\approx 0.7$ is in line with the smallest observed damping ratios ($\\approx 0.6$), suggesting wave leakage is a significant damping agent in real weak-field filaments.","The transition between pendulum-valid and pendulum-invalid behavior is governed by $\\delta$, not by plasma $\\beta$; stronger-field filaments with $\\delta\\approx 0.2$ remain safe for the pendulum model."],"supporting_citations":[{"why":"Establishes the 1D pendulum model in which field-aligned gravity reproduces observed filament longitudinal oscillation periods, the baseline the paper challenges in the delta~1 regime.","marker":"Luna & Karpen (2012)"},{"why":"1D non-adiabatic simulation reproducing an observed period but with decay time 1.5 times too long, the discrepancy that motivates adding wave leakage.","marker":"Zhang et al. (2012)"},{"why":"Defines the dimensionless gravity-to-Lorentz parameter delta = rho g L/(B^2/2 mu0) and argues it, not plasma beta, controls whether filament gravity deforms the magnetic field.","marker":"Zhou et al. (2018)"},{"why":"Earlier 2D MHD simulations with delta~0.2 found only slight field deformation and a valid pendulum model, the contrast case for the new delta~1 simulations.","marker":"Luna et al. (2016)"},{"why":"Shows wave leakage is ineffective for a 3D flux tube, the objection the paper answers with the sheared-core/unsheared-envelope piston argument.","marker":"Terradas et al. (2006)"},{"why":"Compiles observed decay-to-period ratios as small as 0.6, providing the observational target the paper's tau~0.7P approaches.","marker":"Luna et al. (2018)"},{"why":"Describes the sheared core field overlain by an unsheared envelope in solar filaments, the topology used to argue the 2D piston mechanism survives in 3D.","marker":"Chen 2011"}],"fun_headline_variants":["2D MHD: Wave leakage damps filament oscillations","Weak-field filaments leak waves, skew pendulum model","Wave leakage doubles filament seismology error","Filament oscillations shed waves, breaking pendulum model","Pendulum model fails when filament leaks waves"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument relies on the 2D slab result carrying over to the real three-dimensional Sun, where a filament's sheared core field is surrounded by a quasi-perpendicular envelope that forces the ambient field to oscillate like a piston; if the surrounding field lines are instead pushed aside, the extra damping and the 100% pendulum error would shrink substantially.","fun_headline_variants_meta":{"raw":{"variants":["2D MHD: Wave leakage damps filament oscillations","Weak-field filaments leak waves, skew pendulum model","Wave leakage doubles filament seismology error","Filament oscillations shed waves, breaking pendulum model","Pendulum model fails when filament leaks waves"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000271,"raw_usage":{"total_tokens":1666,"prompt_tokens":1019,"completion_tokens":647,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":635,"completion_tokens_details":{"reasoning_tokens":576}},"tokens_in":635,"tokens_out":647,"duration_ms":6799,"temperature":1.0,"reasoning_tokens":576,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:25:03.817661+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A 3D MHD simulation of a filament with $\\delta\\approx 1$ and a realistic sheared-core/unsheared-envelope field would settle it: if the decay time is not roughly half the matched 1D value and no fast-mode wave trains stream away from the filament, the wave-leakage claim fails. Observational support could come from detecting quasi-periodic 810 km/s wavefronts propagating above oscillating filaments in weak-field regions; their absence in many events would indicate the 2D piston picture does not apply.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows wave leakage is ineffective for a 3D flux tube, the objection the paper answers with the sheared-core/unsheared-envelope piston argument."}],"review_version":1}