{"id":"c577887c-5856-4eaa-97fc-91bde624fc3f","arxiv_id":"2411.18771","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In 2.5D and 3D simulations, test protons driven by turbulent Hall-MHD fields gain more energy and spread more widely than protons in a self-consistent hybrid particle-in-cell plasma.","lead":"The paper compares two ways of simulating how protons get energized in turbulent magnetized plasma: test particles that only feel the fields, and a self-consistent model in which particles also create fields. The test-particle method heats particles more and spreads fast particles across the whole domain, while the self-consistent method keeps them confined, so test-particle results should be read with caution.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central comparison conflates test-particle approximation with model physics; a kinetic test-particle control is needed.","rationale":"The reader's weakest_assumption identifies the same core concern: the CHMHD and HPIC arms differ in both self-consistency and underlying model, so the energization gap is not cleanly attributable to the test-particle approximation. My independent reading of the manuscript confirms this. The paper's own energy diagnostics (Figures 2 and 3) show that HPIC has a distinctly different energy partition and a collisionless dissipation channel absent from CHMHD, reinforcing the concern. The proposed kinetic test-particle control is the single most direct check: it isolates the effect of particle feedback while holding the kinetic model fixed. The paper does provide some independent support, such as matched large-scale spectra and the use of the same GHOST framework, but these do not eliminate the physical-model confound. The reader's CONDITIONAL verdict is therefore appropriate, and my stress test does not move it. No code or data is provided to rerun the existing simulations, but the suggested control is computationally feasible in the same framework and would settle the attribution of the central claim.","tokens_in":14277,"tokens_out":2059,"duration_ms":18646,"concrete_test":"Run a third 2.5D simulation: evolve the HPIC fields exactly as in the paper, but also integrate a separate population of test protons that react to those fields without depositing charge or current back onto the grid. Then compare mean energization, perpendicular/parallel temperature, and P(kappa>3) of these kinetic test particles against the existing HPIC self-consistent run and the CHMHD test-particle run. If the kinetic test particles reproduce the CHMHD test-particle overestimation, the test-particle approximation is the cause; if they track the HPIC self-consistent result, the difference is attributable to the CHMHD-vs-HPIC model difference rather than to the test-particle approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the test-particle approximation overestimates energization and suprathermal spread, based on the difference between a CHMHD test-particle run and an HPIC self-consistent run. However, those two arms differ not only in particle feedback but also in the underlying physical model: CHMHD treats ions as a fluid with viscosity, resistivity, and a pi=pe, Ti=Te closure, while HPIC treats ions kinetically with massless adiabatic electrons. Section IV A states that 'any difference should be due to kinetic effects and the test particle approximation,' but no run separates these two sources. A control that uses the same kinetic model with and without feedback is absent. Without it, the reported overestimation factors (about 3x in 2.5D, 6x in 3D) could be dominated by the different field evolution, dissipation, and closures, rather than by the removal of particle self-interaction. The single-realization comparison also leaves magnitudes fragile, but the confound is the load-bearing issue.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a numerical comparison of proton energization in driven turbulence between a 'test-particle' approach, where protons are evolved as test particles in the fields of a compressible Hall magnetohydrodynamic (CHMHD) simulation, and a 'self-consistent' hybrid particle-in-cell (HPIC) simulation in which the same particles supply the ion moments that determine the fields. The authors conduct one 2.5D and one 3D case, initialized from the same CHMHD stationary state. They find that the test particles gain about three times (2.5D) and about six times (3D) more perpendicular and parallel temperature increase than the HPIC ions, and that the kurtosis-based suprathermal fraction saturates over the entire domain for test particles while remaining localized for HPIC particles. They conclude that the test-particle approximation overestimates heating and suprathermal production while retaining some qualitative features such as preferential perpendicular heating and a similar early ballistic energization.","tokens_in":14397,"tokens_out":7060,"duration_ms":62480,"significance":"The question addressed is of practical importance because test-particle studies are widely used to model energetic particle transport and acceleration in turbulent space plasmas. If the reported overestimation is robust, the paper would provide a concrete cautionary result and a quantitative reference for the bias introduced by omitting particle feedback. The manuscript also contributes a detailed energy budget of the HPIC run, including the partition of injected energy among magnetic, bulk, and thermal channels, and a kurtosis-based diagnostic of suprathermal spatial distribution. The use of both 2.5D and 3D geometries, with matched initial conditions and forcing, is a step toward practical guidance on when the test-particle approximation can be trusted. However, because the two arms differ in the underlying physical model as well as in particle feedback, the central attribution of the difference to the test-particle approximation is not yet established.","major_comments":[{"comment":"The statement that 'any difference should be due to kinetic effects and the test particle approximation' is an assertion that the numerical design does not support. The CHMHD and HPIC runs differ not only in particle feedback but also in the ion model (fluid versus kinetic), the dissipation operators (viscosity and resistivity versus collisionless processes), and the closure assumptions (Ti=Te with an adiabatic electron fluid versus kinetic ions with a separate adiabatic electron pressure). The reported overestimation factors, about 3x in Figure 4 and about 6x in Figure 8, could therefore be dominated by these model differences. To isolate the test-particle approximation, the authors need a control such as an HPIC run in which the same particles are evolved without feeding back into the fields, or a CHMHD run with self-consistent particle feedback. Without such a control, the conclusion that the test-particle approximation overestimates heating is not established by the present comparison.","section":"Section IV A, first paragraph"},{"comment":"The reference CHMHD run is statistically stationary (energies fluctuate around constant values), while the HPIC run is not: injected energy is only partially balanced by dissipation, so thermal and magnetic energies increase monotonically throughout the simulation. Consequently, the test particles evolve in a stationary electromagnetic field, whereas the self-consistent particles evolve in a field that is gaining energy. The difference in energization between the two arms may reflect this asymmetry in the global energy balance rather than particle feedback. The authors should either modify the HPIC setup to achieve a statistically stationary state, or discuss how the different injection/dissipation balance affects the interpretation of the comparison.","section":"Sections III C and IV A, Figures 2 and 7"},{"comment":"The 3D simulation is explicitly described as 'mostly qualitative' because of lower resolution and shorter duration, yet the conclusions state that the CHMHD test particles are heated 'approximately twice' as much relative to its 2.5D counterpart and about 6 times more than the HPIC case. These quantitative factors are derived from a single short run and may not be converged with respect to resolution or duration. The 3D numbers should be presented as qualitative trends, or supported by a resolution study, before being reported as quantitative factors in the abstract and conclusions.","section":"Section IV B and Section V, Figures 4 and 8"},{"comment":"Each case is a single realization with no ensemble averages or error bars. The quantitative claims of 3x and 6x overestimation are therefore point estimates from one run. While the qualitative separation between CHMHD and HPIC may be robust, the specific magnitudes are fragile. The authors should provide at least an estimate of statistical uncertainty, for example by dividing the domain into sub-boxes or by performing multiple realizations, before presenting these factors as quantitative results.","section":"All results, Figures 2, 4, 8"}],"minor_comments":[{"comment":"The phrase 'split en two cases' should be 'split into two cases'.","section":"Section III C"},{"comment":"The word 'Prandlt' should be 'Prandtl'.","section":"Section II C"},{"comment":"The word 'colissionless' should be 'collisionless'.","section":"Section IV A"},{"comment":"The caption says 'both parallel (left) and parallel (right)'; the left panel should be 'perpendicular'.","section":"Figure 9 caption"},{"comment":"The verb 'resemblances' should be 'resembles'.","section":"Section IV B"},{"comment":"The notation 'kIdi' is used for the injection wavenumber, but Table I uses 'kmin'; clarify the relation between the injection scale and kmin in the text.","section":"Section V"}],"recommendation":"major_revision","confidential_remarks":"The missing control is a central issue; I would expect the editor to require either a new run or a significant weakening of the causal claim. The paper is otherwise well structured and the diagnostics are clearly presented."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is the direct comparison: the same turbulent CHMHD snapshot seeds both test particles in the continued CHMHD run and self-consistent ions in an HPIC run, done in 2.5D and 3D. That is a genuinely useful setup, and the paper does it cleanly. The diagnostics—mean energization, directional temperatures, kurtosis fields, and velocity PDFs—are appropriate and the qualitative conclusion, that test particles overestimate heating and spread suprathermals too broadly, is consistently supported across both runs. The presentation is clear, and the authors do not oversell what the simulations show in terms of solar-wind predictions.\n\nThe soft spot is the one the stress-test note flags, and it is load-bearing. The CHMHD arm treats ions as a fluid with viscosity, resistivity, and a Ti=Te closure; the HPIC arm treats ions kinetically with massless adiabatic electrons. So the comparison conflates the test-particle approximation with the difference between fluid and kinetic models. Section IV A says \"any difference should be due to kinetic effects and the test particle approximation,\" but that overstates the isolation. The fields themselves evolve differently in the two runs, with different dissipation and closures, so the 3x and 6x overestimation factors are not cleanly attributable to the removal of particle feedback. A control is missing: test particles evolved in the same HPIC fields, or any other design that separates feedback from model physics. Without that, the quantitative factors are suggestive rather than demonstrated.\n\nThere is also the single-realization issue—no ensemble, no error bars—which leaves the magnitudes fragile, especially in the shorter, lower-resolution 3D run. And no code or data are provided, which limits independent checking. These are addressable, and the paper's qualitative conclusions would likely survive better controls, but the central quantitative claim needs that additional work.\n\nWho gets value from this: anyone using test particles in turbulent plasma studies, particularly for solar wind and energetic particle populations. It is a good cautionary benchmark and would spark useful discussion at a reading group. But it needs a serious revision to separate the confound before I would trust the numbers.\n\nMy recommendation: send it to peer review, but with the explicit expectation that the authors either add a kinetic test-particle control, weaken the attribution to the test-particle approximation, or both. The idea is worth engaging; the current form is not the final word.","headline":"The paper gives a useful first side-by-side comparison of test-particle and self-consistent kinetic energization in a shared turbulent state, but its central quantitative claim is weakened because the two arms differ in model physics as well as in particle feedback.","tokens_in":14960,"tokens_out":1969,"would_cite":true,"duration_ms":30525,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Test-particle runs overestimate proton heating in turbulent plasma simulations.","keywords":["turbulent plasma","test particle approximation","hybrid particle-in-cell","particle energization","suprathermal particles","Hall magnetohydrodynamics","collisionless dissipation","proton heating"],"falsifier":"Run a third simulation with the same initial turbulent state but with particle feedback added to the CHMHD fluid equations (a two-way coupled fluid-particle scheme); if the energization gap between it and the test-particle branch disappears, the overestimation is not caused by the test-particle approximation alone.","tokens_in":13983,"feed_emoji":"⚡","tokens_out":9534,"duration_ms":76613,"temperature":0.7,"pith_summary":"The paper asks whether the cheap and widely used test-particle approximation—pushing charged particles through precomputed electromagnetic fields without letting them act back on the fields—can reproduce the energization of protons in a turbulent plasma. By directly comparing a compressible Hall magnetohydrodynamic run with test protons against a hybrid particle-in-cell run with self-consistent protons, starting from identical turbulent initial conditions, the authors find that test particles gain substantially more thermal energy: about three times more in the 2.5D run and roughly six times more in the 3D run. They also find that suprathermal particles fill the entire domain in the test-particle case but remain localized in the self-consistent case. The upshot is that test-particle studies are trustworthy for early energization and for the direction of preferential heating, but they overstate heating rates and the abundance of suprathermal particles.","feed_headline":"Test particles overheat in turbulent plasma runs","feed_subtitle":"Self-consistent runs heat 3–6x less and confine suprathermals; test-particle results are biased high.","key_machinery":"The key machinery is a side-by-side simulation protocol: a single stationary turbulent state produced by the CHMHD model (a compressible Hall MHD system with viscous and resistive dissipation) serves as the common initial condition for both branches. From that state, particles are sampled from the local fluid density and temperature, and the run splits: one branch continues the CHMHD evolution while pushing particles as test particles that do not feed back on the fields; the other branch switches to HPIC, a hybrid particle-in-cell scheme with kinetic protons and massless adiabatic electrons, where particles self-consistently generate the fields. The diagnostic that carries the suprathermal-particle comparison is the kurtosis field $\\kappa_\\ell(x,t) = \\delta M_{\\ell,4}(x,t)/T_\\ell^2(x,t)$, computed from the fourth-order centered moment of the local velocity distribution; $\\kappa_\\ell > 3$ marks a suprathermal tail, and the paper tracks the fraction of grid points with $\\kappa_\\ell > 3$ over time.","core_discovery":"The central claim is that removing the coupling from particles back to the electromagnetic fields—the test-particle approximation—biases particle energization upward in turbulent plasmas with a magnetic guide field. The authors demonstrate this by starting both branches from the same stationary CHMHD turbulent state, then evolving one branch self-consistently with the particles (HPIC) and the other with particles riding as test particles on the CHMHD fields. In both 2.5D and 3D simulations, the test-particle branch reaches a higher mean temperature increase: about three times higher in 2.5D and roughly six times higher in 3D at the end of the runs, after a short period where both agree. The suprathermal analysis shows that in the test-particle case the fraction of grid points with kurtosis greater than 3 reaches one across the whole domain, while in the self-consistent case that fraction stays lower and the suprathermal particles remain confined to specific regions. The authors conclude that test particles capture the early energization and the preferential perpendicular heating, but miss finer phenomena and systematically overestimate energization and the width of the velocity tails.","pith_inferences":["A two-way coupled fluid-particle run (CHMHD with particle feedback) would be needed to separate the effect of removing particle feedback from the effect of treating ions as a fluid; the paper does not include such a run, so part of the gap could be model-dependent rather than approximation-dependent.","Because the HPIC results change little between 2.5D and 3D while the CHMHD results change substantially, self-consistent energization may be controlled by local kinetic structures that are insensitive to dimensionality, whereas test particles respond more to the global field geometry.","The paper's qualitative argument that self-consistent high-energy particles generate fields that arrest their own acceleration suggests a feedback saturation mechanism that could be tested by measuring the correlation between suprathermal particle density and local magnetic-field enhancement in the HPIC runs."],"forward_implications":["Test-particle results are reliable for early energization (up to roughly 0.13 large-eddy turnover times) and for the preferential perpendicular heating, but not for absolute heating rates.","Studies using test particles to estimate cosmic-ray or suprathermal ion production in turbulent plasmas will overestimate those rates, because the same normalized velocity distributions give heavier tails in the test-particle case.","The bias worsens in 3D relative to 2.5D, so conclusions drawn from lower-dimensional test-particle runs may be even less reliable when the injection scale is close to ion scales.","Higher-order velocity statistics (kurtosis and tail weight) are more poorly captured by the test-particle approximation than mean temperature, so suprathermal-population studies need self-consistent kinetics.","Self-consistent particles can develop suprathermal tails in the parallel direction even when the simulation is 2.5D, indicating parallel energization mechanisms that are inaccessible to test particles."],"supporting_citations":[{"why":"Supplies the test-particle method in simulated turbulent fields that the CHMHD branch extends and that this paper compares against.","marker":"25–31"},{"why":"Motivates self-consistent kinetic simulations as a more realistic ground truth for particle feedback.","marker":"35,36"},{"why":"Shows hybrid kinetic-fluid models reproduce ion-scale phenomena, justifying the HPIC simulation as the reference.","marker":"38–40"},{"why":"Provides the GHOST code with the particle module used to run the HPIC simulations.","marker":"41"},{"why":"Provides the GHOST compressible Hall MHD solver used to generate the turbulent fields for the test-particle branch.","marker":"42,43"}],"fun_headline_variants":["Test particles overheat plasma runs","Self-consistent particles cool the bias","Turbulent plasma test particles mislead","Particle back-reaction cuts heating 3-6x"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The comparison assumes that the difference between the CHMHD-plus-test-particle run and the HPIC run is caused by the test-particle approximation itself, but the two runs also differ in physical model—CHMHD treats ions as a single fluid with a pi=pe closure, while HPIC treats ions kinetically—and no run isolates these two sources.","fun_headline_variants_meta":{"raw":{"variants":["Test particles overheat plasma runs","Self-consistent particles cool the bias","Turbulent plasma test particles mislead","Particle back-reaction cuts heating 3-6x"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000152,"raw_usage":{"total_tokens":1229,"prompt_tokens":999,"completion_tokens":230,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":174}},"tokens_in":615,"tokens_out":230,"duration_ms":3300,"temperature":1.0,"reasoning_tokens":174,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T10:53:42.630170+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a third simulation with the same initial turbulent state but with particle feedback added to the CHMHD fluid equations (a two-way coupled fluid-particle scheme); if the energization gap between it and the test-particle branch disappears, the overestimation is not caused by the test-particle approximation alone.","supporting_citations":[],"review_version":1}