{"id":"6d4c020b-f890-4dd1-a7fa-247523fa9b07","arxiv_id":"2607.24823","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"GEANT4 modeling shows the Martian regolith is the dominant source of surface neutrons and gammas, and that 0.11 wt.% hydrogen measurably changes thermal and epithermal albedo neutron spectra.","lead":"This paper uses GEANT4 simulations to model neutrons and gamma rays produced at the Martian surface by cosmic-ray protons, comparing dry regolith with regolith containing 0.11 wt.% hydrogen. It concludes that the regolith, not the atmosphere, dominates the surface radiation field and that small hydrogen abundances leave a clear signature in albedo neutron spectra.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Regolith bulk density 3.01 g/cm3 is a solid-crust value; porous surface regolith (~1.6 g/cm3) would change absolute spectra and H sensitivity.","rationale":"I agree with the reader's identification of the regolith density as the weakest assumption. The central claim comprises two assertions—regolith dominance and hydrogen observability—both of which are sensitive to the number density of nuclei in the top 2 m. The density value chosen is for solid crust, not for the porous surface regolith the paper claims to model. A rerun at a realistic value is a direct, inexpensive test. The lack of statistical uncertainties and the unexplained atmospheric-vs-regolith decomposition are real but secondary; the decomposition only affects the first assertion, and the hydrogen comparison is within the same geometry. Therefore the reader's conditional verdict is appropriate, and no change is needed.","tokens_in":9427,"tokens_out":11482,"duration_ms":124787,"concrete_test":"Rerun the identical GEANT4 setup with regolith bulk density 1.6 g/cm3 (with the same wt% composition and 0.11 wt% H), keeping the 2-m thickness and also a constant-column-mass variant (3.75 m). Recompute Fig. 6 total-vs-atmospheric ratios and Fig. 7 thermal/epithermal/fast albedo spectra with and without H, with the same 10^5 protons and a statistical uncertainty estimate. If the total/atmospheric ratio falls below ~1 in any energy bin, or if the with/without-H difference becomes comparable to Poisson noise, the central claim fails; if both survive, the claim is robust to density.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The simulation (Sec. 2, Fig. 1) uses a single 2-m regolith layer with bulk density 3.01 g/cm3, citing petrological estimates of Martian crustal density [11]. The upper 2 m of Mars is porous regolith/soil, with bulk densities typically 1.5–1.8 g/cm3. Neutron and gamma transport depends on the number density of nuclei, so this choice roughly doubles the macroscopic interaction cross sections per unit depth relative to a realistic regolith. This affects the central claim in two ways. First, the regolith's contribution to the surface neutron/gamma population (Fig. 6) is inflated because each primary proton sees ~2x more target mass in the top 2 m; the 'regolith rather than atmosphere dominates' conclusion could be an artifact of an over-dense target. Second, the hydrogen sensitivity (Fig. 7) depends on the H atom column density; at 0.11 wt% H, a 2-m column at 3.01 g/cm3 contains ~0.66 g/cm2 H, whereas at 1.6 g/cm3 it contains ~0.35 g/cm2. The reported thermal enhancement and epithermal deficit may therefore be quantitatively weaker under realistic conditions. Although the qualitative sign is likely robust because the atmosphere is only ~24 g/cm2, the paper's quantitative spectra and the 'observable in each energy regime' claim need to be re-examined at a realistic bulk density.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses GEANT4 simulations to model the interaction of galactic-cosmic-ray protons with the Martian atmosphere and a 2-m regolith layer, with and without 0.11 wt.% hydrogen in the regolith. The authors compute proton attenuation, generation-depth profiles of secondary neutrons and gammas, and compare atmospherically produced and total surface spectra. They conclude that the regolith dominates the surface neutron and gamma population, and that the chosen hydrogen abundance produces a thermal-neutron excess and epithermal/fast-neutron deficit that is observable in each albedo-neutron energy regime.","tokens_in":9798,"tokens_out":2302,"duration_ms":28091,"significance":"If the quantitative results are robust, the paper provides a useful computational confirmation of the expected thin-atmosphere planetary physics: regolith-generated secondaries dominate the surface radiation field, and trace hydrogen leaves a clear spectral signature in albedo neutrons. The qualitative trends are consistent with established orbital and in-situ observations, and the paper makes its simulation setup reproducible through a linked code repository. The main value is as a benchmark, not as a new physical discovery; the quantitative claims, however, depend on a few modelling choices that need to be checked before the conclusions can be accepted.","major_comments":[{"comment":"The regolith is modeled as a 2-m layer with bulk density 3.01 g/cm3, a value appropriate for solid crust, not for the porous upper regolith/soil that actually occupies the top 2 m. Realistic bulk densities are ~1.5–1.8 g/cm3. Because neutron/gamma transport depends on number density, this choice approximately doubles the macroscopic cross sections and the hydrogen column density relative to a realistic regolith. Concretely, 0.11 wt.% H over 2 m at 3.01 g/cm3 gives ~0.66 g/cm2 of H, while at 1.6 g/cm3 it gives ~0.35 g/cm2. The reported dominance of the regolith over the atmosphere (Fig. 6) and the magnitude of the thermal/epithermal hydrogen signature (Fig. 7) are therefore likely overestimated. The authors should rerun the simulations at a porosity-corrected density or justify why the solid-crust value is appropriate for the upper 2 m.","section":"§2, Table 2, Fig. 1"},{"comment":"No statistical uncertainties are shown anywhere. The simulation uses only 1e5 incident protons. Many high-energy bins in Figs. 6 and 7 contain small integer counts, so the visual separation between 'with H' and 'without H' curves — especially the fast-neutron convergence at higher energies — may not be statistically significant. The claim that hydrogen is 'observable in each energy regime' needs a quantitative significance assessment (e.g., Poisson error bars or a hypothesis test per bin/region). Without this, the central observational claim is not supported beyond a qualitative trend.","section":"§3, Figs. 3–7"},{"comment":"The mechanism by which the 'surface detector' selects albedo (upward-going) neutrons is not described. A 2-mm-thick pseudo-detector placed at the atmosphere–regolith interface will record particles crossing in both directions unless a directional filter is applied. The paper does not state how upward-going albedo neutrons are distinguished from downward-going atmospheric or cascade neutrons. This is essential for interpreting Fig. 7 as an albedo spectrum. Please specify the tracking/direction selection criteria.","section":"§2, Fig. 1; §3, Fig. 7"},{"comment":"The atmosphere is modeled as a single 12-km-thick slab of constant density 2×10−5 g/cm3. While the column depth (~24 g/cm2) is in the right ballpark, a constant-density slab is a coarse approximation to the real exponentially decreasing atmosphere and can distort the depth-dependent development of hadronic cascades, particularly the attenuation of low-energy protons and the production altitudes of secondary neutrons/gammas. A quantitative estimate of the sensitivity of the surface spectra to this simplification would strengthen the paper.","section":"§2, atmospheric model"}],"minor_comments":[{"comment":"Typographical and grammatical errors, e.g., 'Then, The energy spectra' in the abstract and inconsistent capitalization. Also, the term 'PAMELA proton beam' is misleading; PAMELA provides a measured spectrum, not a beam.","section":"Abstract and §1"},{"comment":"The y-axis label 'Counts' is ambiguous. It should be stated whether these are raw counts per bin for the 1e5 simulated primaries, counts per primary, or normalized per unit energy. This is necessary for reproducibility and for comparison with other models.","section":"§3, Figs. 4, 6, 7"},{"comment":"Several references are incomplete or lack page numbers, and the GitHub repository [7] is not described in the text. Please ensure all URLs are accessible and cited in the text.","section":"References"},{"comment":"The paper repeatedly states that the proton transport is 'comparatively insensitive' to hydrogen. Since hydrogen replaces oxide mass, this is expected at the 0.1% level; consider stating this expectation explicitly to avoid over-interpreting a null result.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The density issue is the main technical concern: using a solid-crust density for the upper regolith is a load-bearing assumption that inflates both the regolith-vs-atmosphere dominance and the hydrogen sensitivity. The absence of any statistical treatment is also a gatekeeping issue for a simulation paper claiming a quantitative observable. If the authors rerun at realistic density and add uncertainty quantification, the paper could be acceptable. The novelty is modest — the qualitative physics is well established — but the paper could still be a useful computational reference."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a competent GEANT4 parameter study, but the main qualitative claims were already established by Mars neutron spectroscopy and earlier models. The genuinely new part is the explicit atmosphere-versus-regolith decomposition and the fixed 0.11 wt% H comparison. The paper does that cleanly: clear geometry, standard physics list, voxelized depth profiles, and the code is on GitHub. The thermal-excess/epithermal-deficit signature comes out correctly.\n\nThe soft spots are real, though not fatal. The most important is the regolith density. 3.01 g/cm3 is a solid-crust value, not a value for the upper 2 m of porous regolith, which is typically 1.5–1.8 g/cm3. The stress-test arithmetic bears this out: at 0.11 wt% H, the H column over 2 m is roughly twice as large at 3.01 than at 1.6 g/cm3, so both the regolith dominance and the size of the hydrogen effect are inflated. The sign is almost certainly robust—the atmosphere is only ~24 g/cm2—but the quantitative spectra and the claim that H is 'observable in each energy regime' need to be re-examined at a realistic density. This is not a nitpick; it is the central comparison of the paper.\n\nThe absence of statistical uncertainties is the second issue. With 1e5 incident protons, the counts in the figures likely carry visible Poisson errors, but the paper shows no error bars. That makes it hard to judge whether the hydrogen difference in Fig. 7 is significant at every energy bin, especially at the high-energy tails. A sensitivity run at 1.6 g/cm3 and a second seed would tell us a lot.\n\nMinor: the atmosphere is modeled as a constant-density 12-km slab rather than an exponential profile; the column depth is about right, so this likely has little effect on the conclusions, but it is worth a sentence. The method for separating 'atmospherically produced' neutrons from the total at the surface detector is described briefly; I would want a clearer description of what is recorded before interaction with the regolith.\n\nWho is this for? People working on Mars radiation environments or future surface missions who want a quick reference spectrum and an H-sensitivity comparison. It is not a breakthrough, but it is a legitimate extension of the author's lunar work. It deserves a serious referee—the density issue is fixable, and the paper, after revision, could be a useful reference. I would not desk-reject it.","headline":"A competent GEANT4 parameter study with the expected qualitative answers; the crust-like regolith density and missing error bars are the main quantitative weaknesses.","tokens_in":10247,"tokens_out":2533,"would_cite":false,"duration_ms":26709,"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":"Using GEANT4, this paper claims that the Martian regolith, rather than the atmosphere alone, is the dominant source of the neutron and gamma population at the surface, and that 0.11 wt.% hydrogen is detectable as a spectral shift in all thr","keywords":["albedo neutrons","albedo gamma rays","GCR protons","hydrogen detection","Martian regolith","neutron spectroscopy","GEANT4 simulation","thermal/epithermal/fast neutrons"],"falsifier":"Compare the simulation's predicted thermal-to-epithermal albedo neutron ratio for a regolith with 0.11 wt.% hydrogen against a measured neutron spectrum at a well-characterized Martian site with comparable water-equivalent hydrogen content; if the observed ratio does not exceed the dry-site ratio by roughly the predicted amount, the hydrogen-sensitivity claim fails. A cheaper check: rerun the same geometry with a porosity-corrected regolith density of ~1.5 g/cm3 and see whether the regolith contribution still dwarfs the atmosphere by two orders of magnitude.","tokens_in":9325,"feed_emoji":"🪐","tokens_out":6999,"duration_ms":63010,"temperature":0.7,"pith_summary":"This paper uses GEANT4 simulations to model galactic cosmic-ray protons striking a 12-km Martian atmosphere above a 2-meter regolith, both with and without 0.11 wt.% hydrogen. It aims to show that the regolith—not the air—generates the dominant neutron and gamma population reaching the Martian surface, and that even trace hydrogen leaves a clear mark on albedo neutrons. The simulated spectra reproduce the 'epithermal deficit, thermal excess' pattern that orbital neutron spectrometers use to map subsurface water. If correct, the work supports neutron spectroscopy as a sensitive tool for detecting hydrogen and possibly water ice on Mars, and it clarifies that the surface radiation environment must be modeled from the ground up, not just from the atmosphere. The proton transport and the depth profile of secondary production, by contrast, barely change with hydrogen at this level.","feed_headline":"Regolith, not air, drives Mars surface neutron and gamma spectra","feed_subtitle":"Even 0.11% hydrogen shifts albedo neutrons into the thermal band, a signature orbiters can use to map water ice.","key_machinery":"The central mechanism is the GEANT4 Monte Carlo simulation (FTFP_BERT_HP physics list) of the hadronic and electromagnetic cascade: a planar vertical beam of 10^5 protons sampled from the PAMELA spectrum is injected at the top of a 12-km CO2 atmosphere (density 2×10^-5 g/cm3) overlying a 2-m regolith voxelized into 100 cells of 2 cm thickness, with a surface detector at the regolith top. The simulation separates atmospherically produced particles from total surface-detector records and bins albedo neutrons into thermal (≤1 eV), epithermal (1 eV–1 keV), and fast (>1 keV) energy regimes—this energy binning is what exposes the hydrogen moderation signature.","core_discovery":"On the paper's own terms: using a PAMELA-based 69-bin proton spectrum injected vertically into a modeled CO2-dominated atmosphere over a homogeneous 3.01 g/cm3 regolith, the simulations show that the total neutron and gamma population at the surface detector exceeds the atmospherically produced contribution by roughly two orders of magnitude below a few MeV, establishing the regolith as the primary source of the surface radiation field. Adding 0.11 wt.% hydrogen does not alter proton ranges, energy deposition, or the generation-depth profile (which peaks at ~20–40 cm), yet it measurably increases thermal albedo neutrons while reducing epithermal and low-energy fast albedo neutrons—the signat","pith_inferences":["Because the simulation uses a single homogeneous regolith density of 3.01 g/cm3, the absolute fluxes are likely overestimated; a more realistic porous regolith would probably reduce the regolith-vs-atmosphere dominance factor, although the qualitative hydrogen signature should survive.","The vertical planar proton beam is not an isotropic GCR flux; we expect an isotropic or angular-dependent source would alter atmospheric attenuation and could shift the energy at which atmospheric and regolith contributions converge.","The 0.11 wt.% value is a single point; extending the simulation to a series of lower hydrogen abundances could define the detection threshold of neutron spectroscopy and the minimum water-equivalent hydrogen that is observable.","The voxelized depth profile suggests a testable extension: modeling a buried ice table at various depths would predict how the thermal/epithermal albedo ratio changes with ice burial depth, which could be compared with radar or neutron data from polar craters."],"forward_implications":["If the claim holds, orbital neutron spectrometers can detect subsurface hydrogen at the ~0.1 wt.% level from the thermal/epithermal ratio alone, without needing absolute flux calibration.","Radiation-exposure models for future Mars missions must include regolith-generated secondaries, since they dominate the low-energy neutron and gamma environment at the surface.","Trace hydrogen acts as a moderator that redistributes albedo neutron energies rather than changing total neutron production, so hydrogen mapping is inherently a spectral-shape measurement.","The same simulation strategy can be transported to other thin-atmosphere bodies (e.g., the Moon, asteroids) where a similar hydrogen-sensitivity pattern is expected.","Because proton transport is insensitive to 0.11 wt.% H, any observed change in surface neutron spectra can be attributed to hydrogen in the regolith rather than to altered primary beam behavior."],"fun_headline_variants":["Mars regolith, not air, spawns surface neutron and gamma glow","Trace hydrogen reshapes Mars albedo neutrons, simulations show","0.11% H bends Mars neutron spectra: regolith's fingerprint","Regolith rules Mars radiation: H shifts neutron bands","Mars surface neutron field? Regolith, not atmosphere, dominates"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The simulation treats the upper 2 meters of Mars as a single homogeneous layer of density 3.01 g/cm3, but real Martian regolith is porous and typically much less dense; if the density is wrong, the absolute spectra and the atmosphere-versus-regolith split shift, even though the hydrogen thermalization signature likely persists.","fun_headline_variants_meta":{"raw":{"variants":["Mars regolith, not air, spawns surface neutron and gamma glow","Trace hydrogen reshapes Mars albedo neutrons, simulations show","0.11% H bends Mars neutron spectra: regolith's fingerprint","Regolith rules Mars radiation: H shifts neutron bands","Mars surface neutron field? Regolith, not atmosphere, dominates"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000175,"raw_usage":{"total_tokens":1182,"prompt_tokens":865,"completion_tokens":317,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":609,"completion_tokens_details":{"reasoning_tokens":230}},"tokens_in":609,"tokens_out":317,"duration_ms":3900,"temperature":1.0,"reasoning_tokens":230,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T20:27:12.161948+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the simulation's predicted thermal-to-epithermal albedo neutron ratio for a regolith with 0.11 wt.% hydrogen against a measured neutron spectrum at a well-characterized Martian site with comparable water-equivalent hydrogen content; if the observed ratio does not exceed the dry-site ratio by roughly the predicted amount, the hydrogen-sensitivity claim fails. A cheaper check: rerun the same geometry with a porosity-corrected regolith density of ~1.5 g/cm3 and see whether the regolith contribution still dwarfs the atmosphere by two orders of magnitude.","supporting_citations":[],"review_version":1}