{"id":"a9d2ea10-888f-4135-a832-6c08913cc8a5","arxiv_id":"1908.08765","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A variation-based analysis shows that constraining atmospheric muon flux with precision data can reduce sub-GeV atmospheric neutrino flux uncertainty, most effectively when muons are measured at 3000-5000 m altitude sites.","lead":"This paper shows how precisely measured atmospheric muon fluxes can reduce the uncertainty in predicted low-energy atmospheric neutrino fluxes. The key finding is that the benefit depends strongly on the muon observation site, with high mountain measurements working best.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed <10% neutrino-flux uncertainty bound depends on an uncalibrated random-variation ensemble, and the model's own failure on existing BESS data suggests that ensemble does not represent the real hadronic-model error.","rationale":"I read the paper as a coherent proposal for propagating a stipulated hadronic-interaction uncertainty into atmospheric lepton fluxes and for showing that, under that stipulated uncertainty, high-altitude muon measurements can reduce the neutrino-flux uncertainty. The pseudo-analytic kernel formalism is internally consistent, the correlation analysis is reasonable, and the authors are explicit about their assumptions. However, the quantitative claim in the abstract and conclusions—that muon data can reduce the neutrino-flux error to below about 10%—rests entirely on the ad hoc variation model of Sec. III. The reader's weakest-assumption analysis identifies the same load-bearing premise: independent, normally distributed B-spline deformations with δ = 1. My stress-test adds a concrete empirical red flag: in Sec. V, the same model cannot simultaneously reproduce the BESS low-energy muon data at Tsukuba and Mt. Norikura. That failure indicates the assumed random-fluctuation space may be mis-specified, either in amplitude or in correlation structure. Because ς0 is a conditional variance within the assumed ensemble, the reported numerical bounds are not yet validated as actual prediction uncertainties. This does not invalidate the framework; it means the conditional verdict is appropriate. The proposed concrete test—recomputing ς0 under correlated variations with a data-calibrated δ—would directly show whether the central claim survives a more realistic error model. I agree with the reader that the paper should be accepted only conditionally, pending such calibration.","tokens_in":17077,"tokens_out":6083,"duration_ms":73267,"concrete_test":"Recompute the ς0 curves in Fig. 12 with R_ij in Eq. (8) drawn from a Gaussian process with a smooth correlation length of about one decade in log10 p, calibrating δ to the low-momentum BESS muon residuals rather than fixing δ = 1. If any ς0 curve rises above 0.1 where Fig. 12 reports values below 0.1, the claimed reduction is an artifact of the independence assumption and the conditional verdict should be strengthened toward requiring an empirical calibration of the variation model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical results, especially ς0 < 0.1 in Fig. 12, are not a consequence of the muon data; they are a property of the assumed perturbation ensemble defined in Eqs. (8)-(13). The paper takes the difference between the true interaction and the model to be a set of independent normal coefficients R_ij with δ = 1, i.e., a 50% standard deviation per B-spline grid point. No empirical input fixes δ, and no evidence supports the independence of fluctuations across momentum grid points. This matters because ς0 is a conditional variance: given a restriction on muon-flux variation, how much neutrino variation remains within this specific ensemble. A different covariance structure for the real model error can change ς0 substantially. Section V provides a direct low-energy test: the model fails to reproduce the BESS muon fluxes at Tsukuba and Mt. Norikura simultaneously below 1 GeV. That is evidence that the assumed random fluctuations do not capture the actual model discrepancy in the momentum range most relevant to sub-GeV neutrinos. Until the variation model is calibrated against data or against alternative hadronic interaction models, the quantitative bounds should be read as consequences of the assumed 50%-per-grid-point independent variation, not as measured uncertainties.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a method to quantify and reduce the uncertainty in sub-GeV atmospheric neutrino flux predictions by using precision measurements of the atmospheric muon flux. The authors write a pseudo-analytic expression for the atmospheric lepton flux, identify an ``integral kernel'' in the hadronic-interaction phase space, and model model-to-real differences as random independent Gaussian B-spline deformations of that kernel with a 50% per-grid-point amplitude (δ=1, Δ=0.5 in Eqs. 8-13). They generate 3,000,000 random model variations, impose a constraint on the resulting muon-flux variation (Eq. 16), and measure how much the neutrino-flux variation shrinks. The shrinkage is characterized by σ_shrink and fitted to σ_shrink = sqrt(ς0^2 + (ς1 ε)^2) (Eq. 20), with ς0 interpreted as the muon-independent component of the neutrino-flux uncertainty. They compute ς0 for neutrino energies from 0.1 to 100 GeV at Kamioka using muon kernels at Tsukuba, Mt. Norikura, Hanle, and balloon altitude. The main quantitative result is that for muons observed at Hanle (4500 m) down to 0.3 GeV/c, ς0 < 0.1 for all neutrino kinds in 0.15-10 GeV (vertical) and above 0.3 GeV (horizontal) directions. The paper also compares with existing BESS muon data and discusses the influence of projectile-flux ratio and scattering-angle uncertainties.","tokens_in":17342,"tokens_out":8216,"duration_ms":89184,"significance":"If the quantitative claim is taken at face value, the paper offers a practical path to controlling hadronic-interaction uncertainties in atmospheric neutrino flux calculations below 1 GeV using high-mountain precision muon measurements, which is directly relevant to upcoming long-baseline and atmospheric neutrino experiments. The pseudo-analytic integral-kernel framework is a useful conceptual contribution, and the systematic survey of observation sites is informative. The paper is also honestly self-critical: it explicitly reports that the BESS muon data below 1 GeV cannot be simultaneously reproduced at Tsukuba and Mt. Norikura, and it discusses scattering-angle uncertainty as a residual limitation. However, the absolute scale of the reported uncertainties (ς0 values) is set by the uncalibrated assumption of independent 50%-per-grid-point variations, so the central numerical conclusions are conditional on the assumed ensemble rather than directly validated by data.","major_comments":[{"comment":"The amplitude δ=1 and the independence of the normal random coefficients R_ij are assumed without calibration. Since ΔΦ_ν and ΔΦ_μ both scale linearly with δ, the fitted values of ς0 are proportional to the assumed 50% standard deviation per B-spline grid point. The paper provides no external check of this amplitude or of the independence assumption, for example by comparing the ensemble with the spread among existing hadronic interaction models or by using the BESS data to constrain δ. As a result, the absolute statement ς0<0.1 in Fig. 12 is a property of the assumed variation ensemble, not a measured atmospheric-neutrino uncertainty. The site ranking may be more robust than the absolute values, but the paper should either calibrate the ensemble or explicitly present all quantitative bounds as conditional on the assumed variation model.","section":"Sec. III, Eqs. (8)-(13), and Figs. 7-13"},{"comment":"The paper states that below 1 GeV the model fails to reproduce the BESS muon fluxes at Tsukuba and Mt. Norikura simultaneously. This is exactly the energy range most relevant to the paper's central claim. The variation study in Sec. IV conditions on being able to reconstruct the muon flux to within ε≲0.05, but the actual model cannot do that below 1 GeV. Thus the concluding statement that ``when we can reconstruct the atmospheric muon flux measured by a precision experiment, we can also calculate the atmospheric neutrino flux accurately'' is not demonstrated for the low-energy region where it matters most. The paper should either resolve the low-energy BESS discrepancy, for example by including it as a constraint on the variation ensemble, or explicitly limit the conclusions to models that already pass the low-energy muon data.","section":"Sec. V and Sec. VIII"},{"comment":"Eqs. (17)-(19) are internally inconsistent as written. Eq. 17 is not a normal distribution because the exponent lacks a minus sign and the factor 1/2. Eq. 18 then states that the normalized ratio of constrained to unconstrained distributions is exp[(1/σ_ε^2 - 1/σ_∞^2) x^2], which grows with |x| rather than decaying, so the concentration effect is lost. If corrected to a proper Gaussian ratio, the exponent becomes -(x^2/2)(1/σ_ε^2 - 1/σ_∞^2), which is what the subsequent definition of σ_shrink in Eq. 19 requires. The authors should re-derive these equations carefully and ensure the notation in Eq. 17 matches a normalized Gaussian.","section":"Sec. IV, Eqs. (17)-(19)"}],"minor_comments":[{"comment":"There are numerous typographical errors: ``neuron'' should be ``neutrino'' after Eq. (14), ``path'' should be ``pass'' below Eq. (17), ``M. Norikura'' should be ``Mt. Norikura'' in Fig. 8, and ``tat neutrino energy'' in the Fig. 3 caption should be ``at neutrino energy''.","section":"General"},{"comment":"The right-panel caption repeats ``atmospheric neutrino at Eν = 0.1 GeV and atmospheric muon at Pµ = 0.1 GeV/c''; based on the text in Sec. VII it should be Eν = 0.3 GeV and Pµ = 0.3 GeV/c.","section":"Fig. 15 caption"},{"comment":"The criterion for choosing the muon momentum range (correlation larger than 90% of its maximum) is introduced without justification or sensitivity checks. A brief comment on how the results depend on this threshold would help.","section":"Sec. IV"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a serious and readable contribution, and the site survey is valuable. The main reason for major revision, rather than rejection, is that the central method is internally coherent and the BESS discrepancy is explicitly acknowledged; however, the quantitative uncertainty bounds cannot stand as absolute claims without either calibrating the variation ensemble or substantially softening the conclusions. I would support publication if the authors address the calibration issue and the low-energy data failure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe paper is worth reading if you care about atmospheric neutrino flux systematics below 1 GeV. Its real contribution is to make the old idea of calibrating hadronic interactions with muon data quantitative and site-aware. The integral-kernel correlation framework is clear, the ς0 curves are new, and the demonstration that the muon/neutrino phase-space similarity breaks down below 1 GeV is an important caveat for everyone who has used muon data to tune low-energy neutrino flux predictions.\n\nThe paper does several things well. The internal Monte Carlo variation study is coherent; the fitted σ_shrink formula (Eq. 20) actually describes the generated distributions. The site survey (sea level, 2770 m, 4500 m, balloon) is a concrete prescription for future experiments, and the conclusion that a mountain site around 3000–5000 m is better than balloon altitude is a useful, non-obvious result. The authors are also honest about limitations: Sec. VII explicitly checks projectile flux and scattering angle, and the final section concedes the need for accelerator input to reduce the assumed uncertainty.\n\nThe soft spot is the load-bearing variation ensemble. Equations (8)–(13) assume independent normal deformations with δ=1, i.e. 50% standard deviation per B-spline grid point. Nothing empirical fixes δ or the independence. All the quantitative ς0 values, including the headline <0.1 bound for Hanle, are conditional on that ensemble. The stress-test concern is not a straw man: the paper's own BESS comparison (Sec. V) shows the model fails to reproduce the low-energy muon flux at Tsukuba and Mt. Norikura simultaneously. That is direct evidence that the random ensemble does not span the real model discrepancy in the momentum range most relevant to sub-GeV neutrinos. So the <10% numbers should be read as \"under this specific variation model,\" not as measured uncertainties. The scattering-angle uncertainty in Sec. VII is also not folded into ς0 and could be a separate few-percent effect at low energies, especially near horizontal.\n\nI would not block peer review over this. The framework and the site survey are valuable and the assumptions are stated. But the revision should either calibrate the variation amplitude against alternative hadronic models or explicitly relabel the bounds as conditional.\n\nWho should read it: atmospheric neutrino flux people, experimentalists planning muon instruments, and anyone using Honda et al. flux tables for sub-GeV oscillation analyses. It deserves a serious referee.","headline":"A useful, clearly-presented framework for constraining sub-GeV atmospheric neutrino flux uncertainties with muon measurements, but the headline <10% bounds are conditional on an uncalibrated 50%-per-grid-point variation ensemble.","tokens_in":17903,"tokens_out":2566,"would_cite":true,"duration_ms":26816,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.85.Ry","13.85.Tp","13.35.Bv","14.60E"],"model":"deepseek-v4-flash","headline":"Accurate muon flux measurements, especially from a high mountain, can constrain the hadronic-interaction uncertainty in sub-GeV atmospheric neutrino flux predictions to below about 10 percent.","keywords":["atmospheric neutrino flux","atmospheric muon flux","hadronic interaction uncertainty","integral kernel","B-spline variation","sub-GeV neutrinos","muon calibration","neutrino oscillation"],"falsifier":"Compute $\\varsigma_0$ with correlated or larger random deformations of the kernel (for example, smoothing neighboring B-spline coefficients or doubling $\\delta$) and check whether the high-mountain result still gives $\\varsigma_0 < 0.1$ in $0.15 < E_\\nu < 10$ GeV; if not, the claimed reduction is an artifact of the independence assumption. A complementary check is to calibrate two independent hadronic interaction models against the same 5%-accurate mountain muon data and see whether their sub-GeV neutrino fluxes agree to better than 10%.","tokens_in":16858,"feed_emoji":"🏔️","tokens_out":16037,"duration_ms":137370,"temperature":0.7,"pith_summary":"Sub-GeV atmospheric neutrino fluxes are needed for precision neutrino-oscillation experiments, but hadronic-interaction model error dominates their uncertainty. This paper establishes that a precision measurement of the atmospheric muon flux can quantify and, at the right observation site, largely remove that error: if a calculated model can reconstruct the measured muon flux, the same model's neutrino flux is correspondingly close to the true one. Using millions of random deformations of the hadronic production kernel, the authors show the residual neutrino-flux uncertainty can be brought below about 10% ($\\varsigma_0 < 0.1$) across $0.15 < E_\\nu < 10$ GeV vertically and $E_\\nu > 0.3$ GeV horizontally by muon data from a high mountain near 4500 m, whereas sea-level or balloon-altitude data leave larger irreducible errors below 1 GeV. The paper also identifies hadronic scattering-angle uncertainty as a separate source that muon flux normalization alone cannot remove, and shows it can be attacked through the zenith-angle dependence of high-altitude muon fluxes.","feed_headline":"High-mountain muon data cut neutrino flux error to under 10%","feed_subtitle":"Reconstructing muon fluxes measured at 4,500 m keeps model-driven neutrino uncertainty below 10% across 0.15-10 GeV.","key_machinery":"The load-bearing object is the integral kernel $D(N_{\\mathrm{proj}}, p_N, M_{\\mathrm{born}}, p_M, L, p_L, x)$ of the pseudo-analytic lepton-flux formula, which bundles the meson decay probability, hadronic production probability, production cross section, air density, and projectile flux into a density distribution over the hadronic-interaction phase space $(\\log_{10} p_{\\mathrm{proj}}, \\log_{10} p_{\\mathrm{born}})$. Variations of the interaction model are generated by multiplying this kernel by $1 + \\delta \\sum_{ij} R_{ij} B_i(\\log_{10} p_{\\mathrm{proj}}) B_j(\\log_{10} p_{\\mathrm{born}})$, with $R_{ij}$ independent standard normal random numbers and third-order B-spline basis functions; with $\\delta=1$ this gives a 50% standard deviation per grid point. The method then generates millions of random kernel variations, computes the induced neutrino and muon flux variations, and selects only those whose muon flux variation satisfies $|\\Delta \\Phi_\\mu/\\Phi_\\mu| < \\varepsilon$. The standard deviation of the constrained neutrino variation distribution, $\\sigma_{\\mathrm{shrink}}$, is fitted by $\\sqrt{\\varsigma_0^2 + (\\varsigma_1 \\varepsilon)^2}$, so $\\varsigma_0$ is the neutrino-flux uncertainty irreducible by muon data. This machinery is what converts a muon reconstruction residual into a quantitative neutrino flux error.","core_discovery":"The paper's central claim is that the uncertainty in the predicted atmospheric neutrino flux from hadronic-interaction model error can be read off quantitatively from how well the same model reproduces an accurately measured atmospheric muon flux: when the calculated muon flux can be made to match a precision measurement, the calculated neutrino flux is forced close to the true one. Concretely, using variations of the hadronic interaction model built from random B-spline deformations of the production kernel, the authors find that the spread of neutrino flux variations shrinks as the muon flux variation is constrained, and the irreducible part of that spread is the muon-independent uncertainty $\\varsigma_0$. The value of $\\varsigma_0$ depends strongly on the muon observation site: at sea level it rises steeply below 1 GeV, but for a high-mountain site at 4500 m above sea level, with muons measured down to $P_\\mu \\geq 0.3$ GeV/c, it stays below 0.1 for all neutrino species in $0.15 < E_\\nu < 10$ GeV vertically and $E_\\nu > 0.3$ GeV horizontally. The paper further shows that the hadronic scattering-angle uncertainty is a separate error source, producing roughly a 10% neutrino flux change for a 20% scattering-angle change near the horizontal, and that high-altitude muon zenith-angle data can help constrain it. The concluding claim is therefore that reconstructing an accurately measured atmospheric muon flux is sufficient to calculate the atmospheric neutrino flux accurately in the sub-GeV region.","pith_inferences":["The same random-deformation machinery could be rerun with a prior whose amplitude and correlation are tuned to accelerator data on specific production channels; the paper notes that doing so for kaons would suppress the rise in $\\varsigma_0$ at higher energies, but leaves that rerun to future work.","The site dependence implies a design rule the authors do not state: the best muon calibration site is one where muons suffer little energy loss, so their parent-meson phase space overlaps that of sub-GeV neutrinos, which selects mountains near 3–5 km over both sea level and balloon altitude.","A natural next step, using the paper's own ingredients, is a joint fit of high-altitude muon flux normalization and zenith-angle distribution to separate kernel-magnitude errors (tracked by $\\varsigma_0$) from scattering-angle smearing (tracked by the zenith-angle response)."],"forward_implications":["A precision muon measurement on a 3000–5000 m mountain, covering momenta down to 0.3 GeV/c, would keep the residual hadronic-interaction error of atmospheric neutrino flux below 10% for all neutrino species across $0.15 < E_\\nu < 10$ GeV vertically and $E_\\nu > 0.3$ GeV horizontally.","Sea-level muon data, even extended down to 0.3 GeV/c, leave an irreducible neutrino flux uncertainty that rises steeply below 1 GeV, so they cannot fill the same calibration role.","Balloon-altitude muon data produce larger irreducible neutrino uncertainty than mountain data, making balloon programs a less promising route for this calibration.","Reaching 5% neutrino flux accuracy requires reducing hadronic scattering-angle uncertainty to about 10%, since a 20% scattering-angle change induces roughly 10% neutrino flux change near the horizontal; zenith-angle-dependent muon data from a high mountain can contribute to that reduction.","The framework turns muon-flux reconstruction residuals directly into a neutrino-flux error estimate, replacing the earlier qualitative muon-calibration argument with a quantitative one."],"supporting_citations":[{"why":"Establishes the earlier muon-calibration approach that this paper makes quantitative with random kernel variations.","marker":"[9]"},{"why":"Provides the Monte Carlo simulation and atmospheric flux calculation used to build the integral kernels.","marker":"[15, 16]"},{"why":"Supplies the sea-level muon spectrum used as a reconstruction target in the existing-data check.","marker":"[17]"},{"why":"Supplies the 2770 m mountain muon spectrum used to demonstrate the altitude dependence of the reconstruction check.","marker":"[18]"},{"why":"Provides a higher-energy muon spectrum used to extend the comparison of observed and calculated fluxes.","marker":"[19]"},{"why":"Identifies a 4500 m high-altitude site whose muon kernel yields the smallest irreducible neutrino uncertainty in the survey.","marker":"[14]"},{"why":"Provides balloon-altitude muon data that motivate the balloon comparison study.","marker":"[20]"},{"why":"Documents the horizontal enhancement of the three-dimensional neutrino flux that makes scattering-angle uncertainty relevant.","marker":"[22]"}],"fun_headline_variants":["Mountain muons pin neutrino flux error below 10% in sub-GeV range","High-mountain muon data slash neutrino flux error to under 10%","Reconstruct muons at 4.5 km to lock neutrino flux uncertainty below 10%","Sub-GeV neutrino flux error shrinks to <10% using 4.5 km muon data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative bounds depend on the assumption that the true hadronic-interaction error can be represented by random, independent, 50%-per-grid-point deformations of the production kernel and that these deformations do not change the incoming cosmic-ray flux; if real model errors are correlated, larger, or shift the projectile flux, the reported error reductions do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Mountain muons pin neutrino flux error below 10% in sub-GeV range","High-mountain muon data slash neutrino flux error to under 10%","Reconstruct muons at 4.5 km to lock neutrino flux uncertainty below 10%","Sub-GeV neutrino flux error shrinks to <10% using 4.5 km muon data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000966,"raw_usage":{"total_tokens":4198,"prompt_tokens":1120,"completion_tokens":3078,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":736,"completion_tokens_details":{"reasoning_tokens":2983}},"tokens_in":736,"tokens_out":3078,"duration_ms":22572,"temperature":1.0,"reasoning_tokens":2983,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:30:11.311144+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $\\varsigma_0$ with correlated or larger random deformations of the kernel (for example, smoothing neighboring B-spline coefficients or doubling $\\delta$) and check whether the high-mountain result still gives $\\varsigma_0 < 0.1$ in $0.15 < E_\\nu < 10$ GeV; if not, the claimed reduction is an artifact of the independence assumption. A complementary check is to calibrate two independent hadronic interaction models against the same 5%-accurate mountain muon data and see whether their sub-GeV neutrino fluxes agree to better than 10%.","supporting_citations":[{"cited_title":"Study of cosmic ray interaction model based on atmospheric muons for the neutrino flux calculation","cited_arxiv_id":"astro-ph/0611201","evidence_quote":"Establishes the earlier muon-calibration approach that this paper makes quantitative with random kernel variations."},{"cited_title":"Haino et al","cited_arxiv_id":null,"evidence_quote":"Supplies the sea-level muon spectrum used as a reconstruction target in the existing-data check."},{"cited_title":"Sanuki et al., Phys","cited_arxiv_id":null,"evidence_quote":"Supplies the 2770 m mountain muon spectrum used to demonstrate the altitude dependence of the reconstruction check."},{"cited_title":"Achard et al","cited_arxiv_id":null,"evidence_quote":"Provides a higher-energy muon spectrum used to extend the comparison of observed and calculated fluxes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Identifies a 4500 m high-altitude site whose muon kernel yields the smallest irreducible neutrino uncertainty in the survey."},{"cited_title":"Abe et al","cited_arxiv_id":null,"evidence_quote":"Provides balloon-altitude muon data that motivate the balloon comparison study."},{"cited_title":"Battistoni, A","cited_arxiv_id":null,"evidence_quote":"Documents the horizontal enhancement of the three-dimensional neutrino flux that makes scattering-angle uncertainty relevant."}],"review_version":1}