{"id":"3dbae4d1-e97d-4e7d-a1fa-5dbbb10b2cd2","arxiv_id":"2508.17039","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Two eruptive filaments both deflected toward the side with weaker poloidal magnetic field, suggesting inclined filament ejections can result from asymmetric strapping forces.","lead":"This paper tracks two solar filament eruptions with two spacecraft and finds that the filaments accelerated less when the surrounding poloidal magnetic field was stronger, and they veered toward the side where that field was weaker. The authors propose this asymmetric strapping force as a new explanation for why many filaments erupt at a slant.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The weaker-side association hinges on a hand-patched PFSS boundary; a sensitivity test with unpatched and non-potential extrapolations is needed before the claim is secure.","rationale":"The paper's empirical correlation is not circular: the three-dimensional ejection direction is derived from stereoscopic tie-pointing in Section 3.1, while B_pol is computed independently from a PFSS extrapolation. However, the field side is not observed directly; it is the output of a model whose boundary is manually modified. The manuscript itself acknowledges that the synoptic Br in the active region may be observed three to four days away from eruption time and therefore patches it with a local HMI vector magnetogram taken about half an hour before eruption. That patch is an uncontrolled step in the pipeline: there is no sensitivity analysis, no comparison with the unpatched map, and no error propagation into B_pol. The near-limb location of Event 2 (E77) further degrades the reliability of the vector-magnetogram-derived radial field in the patched region. Because the headline claim is directional, even a small systematic bias in the angular gradient of the model field could flip the inferred weaker side. The proposed test, rerunning with unpatched boundaries, perturbed patch positions, and an independent NLFFF/MHD model, would settle whether the weaker-side preference is a genuine property of the coronal magnetic field or an artifact of the extrapolation setup. Until then, the conclusion is a reasonable working hypothesis rather than an established result, so the reader's CONDITIONAL verdict remains appropriate.","tokens_in":13080,"tokens_out":4682,"duration_ms":52850,"concrete_test":"Recompute the Figure 9 ratio analysis for both events after (i) rerunning the PFSS extrapolation with the original synoptic map without the HMI patch, and with the patch shifted by one or two grid cells; and (ii) repeating the B_pol computation with a nonlinear force-free field extrapolation or a data-driven MHD coronal model using the same boundary data. If the ejection-side/opposite-side ratio is not consistently below unity under both sets of runs for both events, the claim that filaments propagate toward the weaker poloidal field cannot be distinguished from a PFSS boundary artifact.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (abstract and Section 4) is that both filaments ejected toward the side with weaker poloidal field. The poloidal field B_pol is computed exclusively from the PFSS extrapolation in Section 2, using a synoptic HMI Br map patched with a local HMI vector magnetogram taken roughly half an hour before each eruption. This creates two unquantified degrees of freedom. First, PFSS is current-free, so non-potential, sheared fields near the erupting flux rope are absent; the external field Bp,ex in Eq. (1) may plausibly be approximated by a potential field, but the patched boundary makes the extrapolation sensitive to the replacement procedure. Second, Event 2 is at E77, near the limb, where the HMI vector magnetogram's radial-field inversion suffers from large projection effects; the patch used for the 2014 event is therefore least reliable exactly in the region that controls the asymmetry. The paper reports no error bars on B_pol, no comparison of patched versus unpatched boundaries, and no independent model. Because the 'ejection side' versus 'opposite side' ratio in Section 3.3 and Figure 9 compares an independently measured direction against a model-dependent field, a boundary artifact in the PFSS solution could produce the apparent weaker-side preference for both events without any physical asymmetry in the coronal strapping force.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes two eruptive filaments (2011 March 7 from AR 11164 and 2014 February 25 from AR 11990) observed simultaneously by SDO and STEREO. Using the tie-pointing technique and 3D linear fitting, the authors reconstruct the three-dimensional ejection trajectories and accelerations. They compute the poloidal component of the external magnetic field from a PFSS extrapolation whose photospheric boundary is a synoptic HMI magnetogram patched with a local vector magnetogram taken half an hour before each eruption. The poloidal field is then compared with the measured acceleration along the path and with the side of ejection relative to the radial direction. The paper reports that the poloidal field strengthens where the ejection acceleration is suppressed, that the acceleration resumes when the decay index exceeds 1, and that both filaments eject toward the side where the poloidal field is weaker. The authors propose that asymmetric strapping by the background field can steer filament eruptions into non-radial directions.","tokens_in":13254,"tokens_out":4527,"duration_ms":43570,"significance":"If confirmed, the result would add a new, observationally grounded mechanism to the known explanations for non-radial filament/CME ejections: asymmetric confinement by the background poloidal field. The strengths of the paper are that the 3D trajectories are obtained from dual-spacecraft triangulation, the apex positions are measured repeatedly to estimate uncertainty, and the decay-index threshold is a standard theoretical value rather than a parameter fitted to the acceleration data. These features make the reported connections between the field geometry and the kinematics more credible than a purely ad hoc comparison. However, the central claim rests on only two selected events and on a single potential-field model with a hand-adjusted boundary; the quantitative robustness of the weaker-side association is not demonstrated. The paper is honest in its final paragraph that further quantification is needed, but the abstract states the conclusion more categorically than the evidence supports.","major_comments":[{"comment":"The central claim of the paper (abstract and Section 4) is that both filaments eject toward the side where the poloidal field is weaker. This asymmetry is computed from a PFSS extrapolation whose boundary is a synoptic HMI map patched by hand with a local vector magnetogram taken half an hour before each eruption. No sensitivity test is reported: the authors do not compare patched and unpatched boundaries, do not vary the patch size or placement, and do not compare with a non-potential extrapolation. The risk is concrete for Event 2, whose patch is at E77° longitude, where the radial-field component of the HMI vector magnetogram is most affected by projection effects. Because the asymmetry in B_pol controls the ejection-side/opposite-side ratio in Figure 9, a boundary artifact could produce the apparent weaker-side preference for both events without any physical asymmetry in the strapping force. A quantitative robustness test (e.g., unpatched versus patched PFSS runs, varying the patch region, or an NLFFF comparison) is needed before the conclusion is secure.","section":"Section 2, Figures 2-3"},{"comment":"The conclusion that 'eruptive filaments tend to propagate along the side with weaker strapping force' is based on two selected events. With N=2, no statistical significance can be attached to the observed agreement, and the paper should either extend the sample or explicitly limit the claim to these two events. At minimum, the authors should report the uncertainty in the measured ejection direction (obtainable from the five repeated apex measurements) and the uncertainty in the B_pol side ratio shown in Figures 9(c-d), so that the reader can judge how robust the 'weaker side' identification is. The last paragraph of Section 4 appropriately notes that further quantification is needed, but the abstract's wording is more categorical than this limitation admits.","section":"Section 3.3, Figure 9, Abstract"},{"comment":"The claimed correlation between acceleration and poloidal field along the ejection path is qualitative. The acceleration profiles have error bars, but statements that the acceleration 'levels off' when the poloidal field 'strengthens to a certain value' and 'resumes' when the decay index exceeds 1 are made without a quantitative test or uncertainty propagation. A cross-correlation, a regression, or at least an explicit identification of the corresponding features with uncertainties would strengthen this secondary claim. In addition, because B_pol and the decay index are derived from the same PFSS model, this part of the analysis is not an independent confirmation of the physical mechanism.","section":"Section 3.2, Figure 7"}],"minor_comments":[{"comment":"The word 'brightennings' should be 'brightenings'.","section":"Section 4"},{"comment":"The word 'extropolation' should be 'extrapolation'.","section":"Section 2"},{"comment":"The transition 'Variously' is awkward; consider replacing it with 'In contrast' or 'For Event 2'.","section":"Section 3.2"},{"comment":"The sentence 'the final positive value of reje arises because the poloidal field reverses on both sides' is ambiguous and should be rewritten to clarify how a ratio that declines to negative values can later become positive.","section":"Section 3.3, Figure 9"},{"comment":"The abstract uses 'appear to eject towards the side', but Section 4 states the weaker-side conclusion without that hedge; the authors should keep the level of certainty consistent between the abstract and the discussion.","section":"Abstract and Section 4"}],"recommendation":"major_revision","confidential_remarks":"This is a well-posed observational study, but the two-event sample and the single hand-patched PFSS model are currently the main limitations. I would encourage the editor to request a revision that adds a sensitivity analysis of the patched boundary, a non-potential or unpatched comparison, and a more cautious framing of the statistical weight of two events. With those changes, the paper could be acceptable for publication in this journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper offers a genuinely new comparison — it directly pits measured three-dimensional ejection directions of two filaments against the computed poloidal-field asymmetry — and both cases behave as the mechanism predicts. The connection is not circular: the acceleration profiles and the poloidal field are derived independently, and the decay-index threshold is a standard criterion, not a fitted parameter. I think it deserves peer review; the referee should request sensitivity tests on the PFSS boundary, not a rejection.\n\nWhat is solid: the event selection is careful, with supporting evidence for flux ropes (hot channels, hook-shaped brightenings), and the tie-pointing trajectories look credible. The observation that the acceleration stalls when the poloidal field strengthens and recovers when the field decays is a nice, concrete piece of phenomenology, particularly for Event 2. The direct comparison to the side-dependent strapping force goes beyond the decay-index approach of Kang et al. and is a reasonable extension of the current-ring model.\n\nThe soft spots are the usual ones for a two-event study, plus one specific technical concern. Two selected events give you no statistics; the inclination angles are modest (14.8 and 26.4 degrees), and the 'weaker side' ratio in Fig 9 shows some non-monotonic behavior, with a sign reversal for Event 2 that the authors explain but which suggests sensitivity to how the ratio is averaged. The bigger issue is the PFSS extrapolation. The boundary is a synoptic map patched with a local vector magnetogram from half an hour before the eruption. For Event 2, that patch sits at E77, near the limb, where the HMI radial-field inversion is least reliable. The paper reports no error bars on Bpol, no patched-versus-unpatched comparison, and no independent non-potential extrapolation. In principle, a patch artifact could generate an apparent weak-side preference without any real asymmetry in the strapping force. This is not a fatal objection — the mechanism is plausible and the acceleration comparison is independent — but it is exactly the kind of thing a referee should ask to see addressed.\n\nBottom line: a serious referee should engage with this. The paper is honest about its limitations and the central claim is framed as a hypothesis. With a sensitivity analysis and a couple more events, this could become a solid reference for how asymmetric confinement steers eruptions. For now, it's a useful case study.","headline":"Plausible new mechanism for non-radial filament eruptions, well-analyzed but dependent on a hand-patched PFSS boundary that needs stress-testing before the claim is secure.","tokens_in":13835,"tokens_out":3039,"would_cite":true,"duration_ms":29241,"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":"Asymmetric magnetic confinement can steer solar filament eruptions sideways.","keywords":["solar filaments","filament eruptions","poloidal magnetic field","strapping force","non-radial ejection","magnetic flux ropes","PFSS extrapolation","solar eruptive kinematics"],"falsifier":"Recompute the poloidal-field asymmetry for the same two events using a nonlinear force-free field or a time-dependent magnetohydrodynamic simulation anchored to the same photospheric boundary, and check whether the ejection still points toward the weaker side. Alternatively, apply the same measurement to a third well-observed eruptive filament whose three-dimensional trajectory is known: a single event that ejects toward the side with stronger $B_{\\rm pol}$ would contradict the proposed rule.","tokens_in":12828,"feed_emoji":"🌞","tokens_out":3747,"duration_ms":39685,"temperature":0.7,"pith_summary":"This paper tries to establish that asymmetric magnetic confinement can steer solar filament eruptions sideways. Using two eruptive filaments observed simultaneously by two spacecraft, the authors reconstruct each filament's three-dimensional ejection path and compare it with the poloidal component of the overlying coronal magnetic field. They find that in both events the eruption accelerates more slowly where the poloidal field strengthens, resumes accelerating where the decay index crosses 1, and finally travels along the side where the poloidal field, and hence the strapping force, is weaker. If correct, this gives a new explanation for non-radial filament ejections and a possible pre-eruption predictor of ejection direction.","feed_headline":"Filaments eject toward the side with weaker magnetic confinement","feed_subtitle":"Two eruptions tracked in 3D tilt toward the weaker strapping force, explaining non-radial ejection.","key_machinery":"The central quantity is the poloidal component of the overlying magnetic field, computed as $B_{\\rm pol} = e_{\\rm pol} \\cdot B_p$ from a PFSS potential-field extrapolation, where $e_{\\rm pol}$ is perpendicular to both the axial current direction and the ejection direction. Because the axial current is nearly constant under line-tying, $B_{\\rm pol}$ directly tracks the strapping force magnitude. The authors compare this quantity along the measured three-dimensional path and on an interception plane perpendicular to the filament axis at the pre-eruption apex, and match its variation to the measured acceleration profile.","core_discovery":"For both investigated events, the filaments appear to eject toward the side where the poloidal magnetic field is weaker, indicating that eruptive filaments tend to propagate along the side with weaker strapping force. The acceleration of both filaments initially rises, then is suppressed or even declines when the poloidal field strengthens, and rises again once the poloidal field decays and the decay index exceeds about 1. The authors interpret this as direct evidence that the downward strapping force controls not only the speed but also the direction of an eruption: when the overlying field is asymmetric, the flux rope yields on the less-confined side.","pith_inferences":["Beyond the two events, the same method could be applied to a larger sample to calibrate how much field asymmetry is needed to produce a measurable deflection, and whether the required ratio varies with flux-rope height.","The paper compares the poloidal field magnitude rather than the full Lorentz force integrated over the flux-rope cross-section; computing the actual strapping force integral along candidate directions could sharpen the directional prediction.","Both events violate the hemispheric helicity rule, hinting that wrong-helicity flux ropes may be especially prone to this asymmetric-confinement steering; a helicity-stratified sample would test whether the effect is limited to that population."],"forward_implications":["Inclined filament ejections can arise simply from an asymmetric overlying field, so forecasting deflection direction may be possible from pre-eruption magnetograms.","A local strengthening of the poloidal field, for example from an additional overlying loop system, can pause or reverse the acceleration of an erupting flux rope.","The ratio of poloidal field strength on the ejection side to the opposite side could serve as a quantitative predictor of non-radial ejection direction.","Models of coronal mass ejection propagation should include the asymmetric strapping force as a steering agent in addition to reconnection and ambient large-scale structures."],"supporting_citations":[{"why":"Supplies the current-ring force balance with hoop force and strapping force that the whole analysis builds on.","marker":"Kliem & Török 2006"},{"why":"Provides the method for computing the poloidal field component along the ejection path from the projected potential field.","marker":"Guo et al. 2019"},{"why":"Establishes the PFSS extrapolation used to reconstruct the global coronal magnetic field.","marker":"Schatten et al. 1969; Altschuler & Newkirk 1969"},{"why":"Describes the tie-pointing triangulation technique used to reconstruct the three-dimensional filament trajectories.","marker":"Inhester 2006"},{"why":"Defines the decay-index threshold used to interpret when acceleration resumes after poloidal-field suppression.","marker":"Démoulin & Aulanier 2010"},{"why":"Presents the directly comparable earlier study comparing ejection direction with decay-index distribution, which this work contrasts with its poloidal-field comparison.","marker":"Kang et al. 2024"}],"fun_headline_variants":["Solar filaments erupt toward weaker magnetic sides","Poloidal field steers solar eruptions to weak side","Erupting filaments bend to weak poloidal field","Filament eruption path set by magnetic asymmetry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison assumes that the extrapolated potential-field model reproduces the real coronal magnetic field, including the patched active-region magnetogram taken half an hour before each eruption; if the actual field is significantly non-potential or the patch distorts the local configuration, the computed weaker side could be an artifact rather than the physical strapping-force asymmetry.","fun_headline_variants_meta":{"raw":{"variants":["Solar filaments erupt toward weaker magnetic sides","Poloidal field steers solar eruptions to weak side","Erupting filaments bend to weak poloidal field","Filament eruption path set by magnetic asymmetry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000469,"raw_usage":{"total_tokens":2325,"prompt_tokens":921,"completion_tokens":1404,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":537,"completion_tokens_details":{"reasoning_tokens":1346}},"tokens_in":537,"tokens_out":1404,"duration_ms":10658,"temperature":1.0,"reasoning_tokens":1346,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T17:07:43.091675+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the poloidal-field asymmetry for the same two events using a nonlinear force-free field or a time-dependent magnetohydrodynamic simulation anchored to the same photospheric boundary, and check whether the ejection still points toward the weaker side. Alternatively, apply the same measurement to a third well-observed eruptive filament whose three-dimensional trajectory is known: a single event that ejects toward the side with stronger $B_{\\rm pol}$ would contradict the proposed rule.","supporting_citations":[],"review_version":2}