{"id":"6013928a-1e99-4df8-97d6-4f0db3a1779c","arxiv_id":"2411.09143","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"GRMHD simulations of magnetically arrested disks reveal a transient stagnation surface near the black hole during flux eruptions, proposed as a site of charged-particle acceleration.","lead":"Simulations of black hole accretion disks show a transient 'stagnation surface' near the event horizon during magnetic flux eruptions, where plasma flow diverges and is depleted. The authors propose this surface could accelerate charged particles, estimating a voltage of about 10^16 volts for M87, as a complementary mechanism to magnetic reconnection.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The stagnation surface is identified by the same density floor that ideal GRMHD uses to erase the vacuum gaps on which the E∥ acceleration claim depends; neither physical existence nor acceleration is established, and Eq. (10)'s voltage estimate is internally inconsistent.","rationale":"The paper reports an interesting morphological feature in 2D GRMHD MAD simulations during flux eruptions, and the use of BHAC plus two identified events is a useful starting point. However, the headline claim is not just that a stagnation surface exists, but that it 'can be associated with an accelerator of charged particles.' To make that claim true, one needs both a physical surface and a credible acceleration mechanism. The manuscript's evidence for the surface is the floor-activation diagnostic; the same diagnostic is a numerical source that operates exactly where ideal GRMHD breaks down. Because the floor adds mass and momentum in those cells, the divergent velocity shown in Figs. 4–5 is not independent of the tracer. This is the weakest load-bearing premise. The reader flagged the floor-artifact possibility; I partially agree, but I would put the emphasis on the floor's active role in the detection rather than primarily on 2D geometry. The numerical energy estimate is independently inconsistent (Eq. 10 vs the 4.8 MeV text and the Pstag value), so there is no reliable quantitative support for the accelerator. A floor-sensitivity study would settle the physical-existence question; if the feature is robust, the paper could be salvaged as a morphological finding, but the acceleration claim would still require kinetic simulations or a non-ideal treatment. Thus the reader's rejection is justified, and my stress test does not change the verdict.","tokens_in":7724,"tokens_out":6818,"duration_ms":70931,"concrete_test":"Rerun the fiducial MAD setup (same initial data, resolution, spin, and final time) with σmax set to 100, 300, and 3000, and if possible with an alternative floor prescription that does not rescale the velocity, then compare the stagnation surface traced by Madd/Ṅadd (location, length, orientation from Eqs. 8–9) and the velocity-divergence pattern across runs. If the feature is invariant under an order-of-magnitude change in σmax, the floor is not constitutive; if it moves, weakens, or disappears, the reported surface is a numerical artifact and the acceleration claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires (1) a physically real stagnation surface and (2) an unscreened parallel electric field along it. The 2D ideal-GRMHD simulation cannot supply (2): ideal MHD enforces E·B=0, and the density floor (§3.2, Eq. 2) injects mass whenever ρ < b²/σmax, actively preventing the low-density 'gap' from which E∥ is inferred. The floor is not a passive tracer: it modifies the primitive state in exactly the cells used to define the surface. Thus the strong spatial coincidence between velocity divergence and Madd/Ṅadd (Figs. 4–5) is evidence about where the numerical floor fires, not independent evidence for a physical stagnation surface. This circularity is the load-bearing weakness: if the floor creates or strongly reshapes the divergent flow, the central 'accelerator' conclusion has no subject. The quantitative support is additionally broken: Eq. (10) gives ΔV~3×10^16 V, yet the text converts this to '4.8 MeV per particle' and then computes Pstag=4.8×10^46 erg/s as if using 3×10^16 eV; the units are off by many orders of magnitude, so the only numerical acceleration estimate is unreliable. The paper's own §5 concedes that 'a more accurate representation' requires 3D simulations; that is a limitation, but the deeper issue is that the 2D floor-based diagnostic cannot distinguish a physical stagnation surface from a numerical artifact.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies two-dimensional axisymmetric GRMHD simulations of a magnetically arrested disk (MAD) around a spinning black hole and searches for regions of divergent velocity during magnetic flux eruption events. The authors introduce cumulative diagnostics of the density-floor activation rate and of the mass added by the floor, and report a transient, near-radial stagnation surface at roughly 2--3 gravitational radii, with length 7--9 gravitational radii, seen in two separate eruption events. They then estimate a potential difference of about 10^16 V along this surface for M87 and propose that such stagnation surfaces act as particle accelerators, complementing or replacing magnetic reconnection as an acceleration mechanism.","tokens_in":8040,"tokens_out":3500,"duration_ms":43382,"significance":"If the stagnation surface were established as a physical feature and the parallel electric-field argument were valid, the result would be a potentially interesting new site for particle acceleration in MADs. The paper's strengths are the clear presentation of a new simulation diagnostic (cumulative floor activation and mass addition) and the demonstration that two flux-eruption events show coincident velocity divergence and enhanced floor activity. These are useful observational tracers within the simulation. However, the central acceleration claim is not supported by the evidence presented: ideal GRMHD enforces E·B=0, the density floor actively reshapes the cells used to identify the surface, and the only quantitative voltage/power estimate contains a unit inconsistency. As it stands, the paper establishes a numerical coincidence, not a physical accelerator.","major_comments":[{"comment":"The acceleration estimate is not derived from the simulations. The simulations solve ideal GRMHD, which enforces E·B=0 in the evolved plasma, so no parallel electric field exists in the simulated state. The text assumes, rather than demonstrates, that 'the electric field reaches values close to the magnetic field strength.' This assumption is load-bearing because the entire conclusion that the stagnation surface is an accelerator rests on the existence of an unscreened E∥. The density floor of Eq. (2) actively fills cells with ρ < b²/σmax, so the low-density 'gap' from which E∥ is inferred is not a physical vacuum gap but a region where the numerical floor is supplying plasma. The paper needs either a non-ideal treatment that produces E∥ or a clearly labeled speculative model, not a statement that the simulated surface 'can be associated with an accelerator.'","section":"§4.2, Eq. (10)"},{"comment":"The voltage and power estimates are internally inconsistent. Eq. (10) gives ΔV ≈ 3×10^16 V; multiplying by the elementary charge gives 3×10^16 eV, which is 3×10^10 MeV, not 4.8 MeV. The stated 4.8 MeV corresponds to a potential of 4.8 MV. The total power Pstag ≈ 4.8×10^46 erg/s is consistent with 3×10^16 eV per particle times 10^42 s^-1 (since 3×10^16 eV ≈ 4.8×10^4 erg), but it is not consistent with the stated 4.8 MeV per particle. Because these numbers form the only quantitative support for the acceleration claim, the arithmetic error invalidates the specific acceleration estimate until corrected and recomputed.","section":"§4.2, Eq. (10) and following text"},{"comment":"The floor-activation diagnostic is not independent evidence for a physical stagnation surface. The floor routine modifies the primitive state by adding mass in precisely the cells where σ exceeds σmax, and the paper identifies the stagnation surface using the resulting floor pattern. The coincidence of divergent velocity with enhanced Madd and Ṅadd therefore shows where the numerical floor acts, but it does not by itself prove that a physically real depletion region exists. The authors' own caveat in §5 that 'a more accurate representation of this phenomenon would require 3D simulations' and the relatively low σmax=1000 make this concern concrete. This issue is load-bearing because if the floor creates or strongly reshapes the divergent flow, the claimed 'accelerator' has no physical subject.","section":"§3.2 and §4.2, Figs. 4–5"}],"minor_comments":[{"comment":"There are several typographical errors: 'relativstic processes' should be 'relativistic processes', 'stars:neutron' should be 'stars: neutron', and 's upermassive' appears as 's upermassive' in the text.","section":"Abstract and key words"},{"comment":"The vector potential formula in Eq. (1) is typeset ambiguously: the expression 'max[...]' with an embedded '− 0.2,, 0' should be cleaned up so the reader can see the functional form being maximized.","section":"§3.1, Eq. (1)"},{"comment":"The expression for γIC mixes several rescaled quantities without clear grouping; in particular, R(h/rs)^{1/2} is ambiguous about which quantities are multiplied. Please define h and rs and parenthesize the factors.","section":"§4.2, Eq. (12)"},{"comment":"The caption's phrase 'Top right and left panel' is confusing because the figure has four panels; please describe each panel explicitly.","section":"Fig. 2 caption"}],"recommendation":"reject","confidential_remarks":"The manuscript reports an interesting numerical coincidence, but the central physical claim is not supported by ideal-GRMHD simulations, and the quantitative acceleration estimate contains a unit error. These are load-bearing issues that cannot be repaired within the current manuscript's scope; a substantially different approach (e.g., force-free or PIC treatment of the gap, or a careful non-ideal model) would be needed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is real: a transient, near-radial stagnation surface that forms during flux eruptions in 2D GRMHD MAD simulations, located at 2–3 r_g and extending 7–9 r_g. The authors track it with a floor-activation diagnostic, and the velocity divergence in Fig. 4 is convincing. That feature is not in the cited literature, and the periodic appearance across two events makes it look like a genuine simulation phenomenon, not a one-off artifact. Credit is due for spotting it and for being explicit about the limitations of 2D and the σmax threshold.\n\nThe soft spots are concentrated in the acceleration claim. Ideal GRMHD enforces E·B=0, so the parallel electric field that the paper needs is assumed, not derived. The floor routine that defines the diagnostic also fills the low-density regions where a physical gap would form, so the spatial coincidence between divergence and floor activation is not independent evidence for a real accelerator. The voltage estimate is also internally inconsistent: Eq. (10) gives ΔV ≈ 3×10^16 V, which is 3×10^16 eV per particle, yet the text says 4.8 MeV; the Pstag calculation uses the correct 3×10^16 eV, so it is a typo, but in a Letter it matters. More fundamentally, the step from a divergent velocity in ideal MHD to a particle accelerator that produces TeV flares requires a kinetic treatment, or at least a force-free/GRMHD hybrid that allows E∥. The paper's own conclusion concedes that 3D simulations are needed, which is honest but does not fix the physics gap.\n\nNone of this kills the observation. The stagnation surface itself is a useful finding for anyone modeling MAD dynamics, and the floor-activation diagnostic may help identify such surfaces in other simulations. The acceleration mechanism is plausible but unproven, and the paper should be read as a simulation-discovery letter with a speculative astrophysical implication.\n\nRecommendation: send it to peer review. A serious referee can separate the solid new feature from the unsupported acceleration claim and push the authors to either reframe the conclusion or add supporting arguments. I would not desk-reject it, but I would not accept it as is.","headline":"A genuinely new simulation feature—a transient near-radial stagnation surface during MAD flux eruptions—is reported well, but the particle-acceleration conclusion is not supported by ideal GRMHD physics and the voltage estimate has a unit error.","tokens_in":8575,"tokens_out":2120,"would_cite":true,"duration_ms":96183,"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 simulations of magnetically arrested disks, a transient stagnation surface forms during flux eruptions and accelerates particles.","keywords":["magnetically arrested disk","flux eruption","stagnation surface","particle acceleration","GRMHD simulation","M87","density floor","black hole jets"],"falsifier":"A three-dimensional general-relativistic magnetohydrodynamic simulation of the same equilibrium torus and spin, with a higher magnetization threshold or a physical pair-production prescription, should reproduce the same near-radial stagnation surface at 2-3 gravitational radii during flux eruptions; if the divergent velocity and floor activation disappear or lose their correlation, the reported surface is an artifact of the 2D geometry or the density floor.","tokens_in":7509,"feed_emoji":"⚡","tokens_out":9732,"duration_ms":87645,"temperature":0.7,"pith_summary":"During magnetic flux eruptions in a simulated magnetically arrested disk (MAD), the authors identify a thin, transient surface near the event horizon where plasma flows diverge and matter is rapidly depleted. They argue this 'transient stagnation surface,' located at about 2-3 gravitational radii and extending 7-9 gravitational radii, is a genuine physical feature rather than merely a numerical artifact. For a supermassive black hole like M87, they estimate a potential difference of roughly $10^{16}$ volts along the surface, enough to accelerate charged particles to very high energies. If true, this gives a complementary or alternative mechanism to magnetic reconnection for producing high-energy and TeV-scale emission from black-hole jets.","feed_headline":"Flux eruptions reveal a 10^16-volt accelerator near M87","feed_subtitle":"Simulations find a transient acceleration layer in magnetically arrested disks that rivals magnetic reconnection.","key_machinery":"The central object is the transient stagnation surface: a near-radial thin layer where the velocity field diverges so plasma streams away in opposite directions, rapidly emptying the region. Because ideal GRMHD cannot represent the evacuated ultra-magnetized plasma, the simulation's density floor must refill those cells; the authors use the cumulative floor activation rate as a diagnostic that traces the surface even where density and magnetization maps look smooth. They combine this with a potential-difference estimate based on the electric field approaching the magnetic field strength over a gap length, following the standard magnetospheric-gap picture.","core_discovery":"The paper reports that in two-dimensional axisymmetric general-relativistic magnetohydrodynamic simulations of a MAD, each flux eruption is accompanied by a transient, near-radial stagnation surface at 2-3 gravitational radii from the black hole, with length 7-9 gravitational radii, defined by a divergent velocity field and traced by enhanced activation of the density floor. The authors show velocity streamlines diverge exactly where the floor injects mass, and a discontinuity in the perpendicular velocity component sits at this surface. They interpret this as a local depletion of plasma that enables a strong electric field parallel to the magnetic field, producing an estimated potential drop of about $10^{16}$ volts for M87, with particle energies capped by synchrotron and inverse-Compton cooling; synchrotron emission would peak in the MeV range and inverse-Compton in the TeV range. The claim is that this is a complementary or alternative particle-acceleration channel to magnetic reconnection during flux eruptions.","pith_inferences":["The same floor-activation diagnostic could be applied to three-dimensional simulations, where non-axisymmetric instabilities may split the single surface into multiple or fragmented acceleration layers.","The voltage estimate scales linearly with the assumed gap length; choosing a different fraction of a gravitational radius would shift the predicted acceleration power by orders of magnitude, so the mechanism's observable brightness is highly sensitive to that assumption.","If the surface is physical, its near-radial orientation means line-of-sight effects should make the resulting MeV/TeV emission strongly dependent on viewing angle and on which side of the jet the eruption occurs."],"forward_implications":["At least two flux eruptions in the simulation produce the same transient stagnation surface about 2-3 gravitational radii from the black hole, extending 7-9 gravitational radii.","The estimated potential difference of about $10^{16}$ volts along the surface for M87-like parameters is large enough to accelerate charged particles to ultrarelativistic energies.","Because the accelerated particles cool through synchrotron and inverse-Compton emission, the model predicts MeV-scale synchrotron and possibly TeV-scale inverse-Compton radiation during flux eruptions.","This provides a complementary or alternative channel to magnetic reconnection for powering high-energy emission from black-hole jets."],"supporting_citations":[{"why":"Supplies the magnetospheric-gap picture and the inverse-Compton cooling cap used for the voltage and maximum Lorentz factor estimates.","marker":"Levinson & Rieger 2011"},{"why":"Establishes the MAD state with large flux eruptions and provides the vector potential and initial setup that produces a MAD in the simulation.","marker":"Tchekhovskoy et al. 2011"},{"why":"Describes how magnetic flux blobs move outward during eruptions because of magnetic tension, motivating the funnel-boundary motion that creates the divergence.","marker":"Dexter et al. 2020"},{"why":"Shows that during eruptions the funnel becomes cylindrical and magnetic reconnection occurs, the background picture this mechanism complements.","marker":"Ripperda et al. 2022"},{"why":"Introduced magnetically arrested disks, the accretion state this paper simulates.","marker":"Narayan et al. 2003"},{"why":"Supplies the M87 magnetic-field strength range used in the voltage and radiation estimates.","marker":"Event Horizon Telescope Collaboration et al. 2019"},{"why":"Provides the numerical code and methods used for the GRMHD simulations.","marker":"Porth et al. 2017"},{"why":"Supplies the density-floor technique and the notion of a continuous inflow-outflow stagnation surface that this paper distinguishes from its transient surface.","marker":"McKinney 2006"}],"fun_headline_variants":["Stagnation surface near M87 yields 10^16-volt drop","MAD eruption surface accelerates particles to TeV","New particle accelerator: stagnation surface in MADs","M87's transient surface rivals reconnection for particle boost","Flux eruptions create 10^16-V potential at black hole"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the divergent velocity and coincident floor activation trace a real plasma-depletion surface in the magnetized funnel, not a numerical artifact of the simulation's matter-refill routine in a 2D axisymmetric run.","fun_headline_variants_meta":{"raw":{"variants":["Stagnation surface near M87 yields 10^16-volt drop","MAD eruption surface accelerates particles to TeV","New particle accelerator: stagnation surface in MADs","M87's transient surface rivals reconnection for particle boost","Flux eruptions create 10^16-V potential at black hole"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000321,"raw_usage":{"total_tokens":1835,"prompt_tokens":1000,"completion_tokens":835,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":616,"completion_tokens_details":{"reasoning_tokens":752}},"tokens_in":616,"tokens_out":835,"duration_ms":12385,"temperature":1.0,"reasoning_tokens":752,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T20:59:27.453330+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A three-dimensional general-relativistic magnetohydrodynamic simulation of the same equilibrium torus and spin, with a higher magnetization threshold or a physical pair-production prescription, should reproduce the same near-radial stagnation surface at 2-3 gravitational radii during flux eruptions; if the divergent velocity and floor activation disappear or lose their correlation, the reported surface is an artifact of the 2D geometry or the density floor.","supporting_citations":[{"cited_title":"& Rieger, F","cited_arxiv_id":null,"evidence_quote":"Supplies the magnetospheric-gap picture and the inverse-Compton cooling cap used for the voltage and maximum Lorentz factor estimates."},{"cited_title":"2020, MNRAS, 497, 4999 Event Horizon Telescope Collaboration, Akiyama, K., Alberdi, A., et al","cited_arxiv_id":null,"evidence_quote":"Describes how magnetic flux blobs move outward during eruptions because of magnetic tension, motivating the funnel-boundary motion that creates the divergence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the density-floor technique and the notion of a continuous inflow-outflow stagnation surface that this paper distinguishes from its transient surface."}],"review_version":1}