{"id":"87a3fcc1-792b-4b15-b932-0f2917dc510a","arxiv_id":"2505.09111","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A sensitivity study showing Hyper-Kamiokande can use atmospheric neutrinos to reconstruct the primary cosmic ray spectrum, cutting its flux uncertainty from about 20% to 7% and sharpening neutrino oscillation measurements.","lead":"This paper simulates how the future Hyper-Kamiokande neutrino detector could measure the cosmic ray spectrum by catching neutrinos produced when cosmic rays hit the atmosphere. If the projection holds, neutrino detectors become a complementary full-sky cosmic ray observatory and improve precision on neutrino mixing parameters.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Hadronic-interaction uncertainties are shown in Appendix A but omitted from the test statistic behind the headline model-discrimination and reconstruction claims, despite the abstract saying they are accounted for.","rationale":"The reader's verdict is CONDITIONAL, with the weakest assumption identified as the proton-only treatment of helium. I agree that is a real gap: the reconstructed 'primary spectrum' in Figure 5 is effectively a proton template, and the 10–20% helium contribution is not propagated into the quoted 5–10% flux uncertainties. However, the most load-bearing issue for the central claims is the mismatch between the abstract and the actual test statistic regarding hadronic interactions. Section IV's q includes only normalization and a 10% cross-section nuisance; the Figure 3 caption says the interaction model is held fixed. The hadronic-model bands are confined to Appendix A. Since the weakest discrimination claim (HKKM New vs AMS) is around 1.1–1.4 sigma in the 1–10 GeV window, adding a ~5% hadronic systematic can plausibly remove it. The same omission inflates the apparent reconstruction precision. This is not a disagreement with current consensus; it is an internal inconsistency between the stated accounting for hadronic uncertainties and the statistic that produces the headline numbers. It is addressable by re-running the analysis with the hadronic spread included, so CONDITIONAL remains the right verdict.","tokens_in":12956,"tokens_out":9544,"duration_ms":104188,"concrete_test":"Recompute the Section IV q statistic for HKKM New and PAMELA relative to AMS with the four MCEq hadronic models included as a discrete nuisance or as a covariance matrix derived from Appendix A, and repeat the Section V template fit with that nuisance included. If the HKKM New significance drops below 1 sigma or the reconstructed-spectrum uncertainties exceed about 10%, the headline claims that hadronic uncertainties are accounted for and that flux uncertainty is reduced to ~7% are not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract states that the models can be differentiated with high significance 'even when accounting for uncertainties in neutrino cross sections and hadronic interactions,' but the main test statistic does not include hadronic-interaction uncertainties. In Section IV, q is maximized only over a free normalization and a 10% Gaussian cross-section nuisance, and Figure 3's caption explicitly says 'holding the interaction model fixed.' The spread from SIBYLL 2.3, EPOS-LHC, QGSJET-II, and DPMJET-III is shown only in Appendix A as count-ratio bands, not propagated into q or into the Section V template-fit uncertainties. This matters most for the smallest claimed discrimination: HKKM New vs AMS is only 1.1–1.4 sigma when the energy range is restricted to 1–10 GeV, where Ref. [75] places hadronic-model uncertainty at about 5%. Adding this as a correlated shape nuisance can plausibly erase that signal. The same omission affects the reconstructed-spectrum precision of 5–10% in Section V, so the quoted reduction to ~7% is not the total uncertainty. The proton-only helium assumption in Section II is a separate idealization; it may be conservative, but it also needs propagation before the reconstruction claim can be taken at face value.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes that next-generation kiloton-scale neutrino detectors, using Hyper-Kamiokande as a concrete example, can measure the primary cosmic ray spectrum through atmospheric neutrinos. The authors simulate 10 years of CC νe and νμ events, compare neutrino spectra produced by injecting different primary CR models (AMS, PAMELA, HKKM Old/New), and define a maximum-likelihood test statistic q with a free normalization and a 10% cross-section nuisance. They report model-discrimination sensitivities from 1.1σ to over 5σ. They then reconstruct the primary CR spectrum by fitting seven single-proton-line neutrino templates to the pseudo-data, claiming a reduction in flux uncertainty from roughly 20% to about 7%. Finally, they use the reconstructed spectra in a simplified sin²θ23 analysis, reporting a factor-of-2 improvement in sensitivity when the CR flux uncertainty is reduced from 15% to 7%. The paper closes with an outlook for JUNO, DUNE, and other detectors.","tokens_in":13179,"tokens_out":3919,"duration_ms":40408,"significance":"If the quantitative claims are correct, this is a useful and timely demonstration that atmospheric neutrino detectors can serve as complementary cosmic-ray monitors, with full-sky coverage and long-term stability, and that improved CR knowledge can sharpen oscillation measurements. The paper's strengths include a coherent closed-loop simulation pipeline (MCEq with multiple hadronic models, HKKM11 plus low-energy FLUKA/CORSIKA fluxes, GENIE cross sections, nuCraft oscillograms), a meaningful cross-check of the injected models against external AMS/PAMELA data, and clear figures that show the energy-dependent nature of the discriminating power. The central idea is credible and the analysis is reproducible in structure. The main weaknesses are that the headline test statistic and the template reconstruction do not actually propagate the hadronic-interaction uncertainties that the abstract claims are accounted for, and the helium-composition approximation is asserted rather than demonstrated. These issues are fixable within the scope of the paper, but they affect the central quantitative claims.","major_comments":[{"comment":"The abstract states that the models can be differentiated 'even when accounting for uncertainties in neutrino cross sections and hadronic interactions,' but the test statistic q in Section IV is maximized only over an overall normalization and a 10% Gaussian cross-section nuisance, and Figure 3's caption explicitly says 'holding the interaction model fixed.' The hadronic-model spread shown in Appendix A (SIBYLL 2.3, EPOS-LHC, QGSJET-II, DPMJET-III) is not propagated into q. This matters most for the smallest claimed discrimination: HKKM New versus AMS is only 1.1–1.4σ at Eν ∈ [1,10] GeV, where Ref. [75] places hadronic-model uncertainty at about 5%. A correlated shape nuisance from the hadronic-model spread could plausibly erase that signal. The authors should either propagate the Appendix A bands through q or revise the abstract and Section IV claims to state that the significances are conditional on a fixed hadronic interaction model.","section":"Abstract and Section IV (Fig. 3)"},{"comment":"The claimed reduction of the neutrino flux uncertainty from ~20% to ~7% is obtained from a closure test: the pseudo-data and the seven template spectra are both generated with the same MCEq/HKKM machinery, and the fit includes only a 10% cross-section nuisance and a free normalization. This measures how well the seven-parameter template model can be constrained internally, not the absolute uncertainty of the atmospheric neutrino flux. The comparison to Super-K's 14.3% and 7.8% flux uncertainties (Ref. [90]) is therefore not apples-to-apples. The paper should explicitly label the 5–10% result as model-conditional and should add hadronic-model and composition uncertainties before claiming a reduction to ~7% total uncertainty.","section":"Section V and Figure 5"},{"comment":"The assumption that the neutrino spectrum produced by helium cosmic rays is identical to that from protons, with the statement that differences 'do not affect our results,' is not supported by any quantitative test. Helium is 10–20% of the primary flux, and the paper itself notes in Section II that shower observables differ between proton- and helium-initiated showers. If helium produces a different neutrino spectral shape, the reconstructed primary spectrum in Figure 5 and the quoted 5–10% uncertainties could be biased. A dedicated MCEq comparison of proton-only versus proton+helium injections, or the inclusion of a helium-shape nuisance parameter in the Section V fit, is needed before the reconstruction claim can be taken at face value.","section":"Section II and Figure 5"}],"minor_comments":[{"comment":"The name 'Sybill 2.3' appears in the text, but the standard spelling is 'Sibyll 2.3' as used in Refs. [68,69].","section":"Section III"},{"comment":"The sentence 'This implies that Hyper-K's statistical uncertainty is below 1%' refers to the total event count, but the energy-binned shape uncertainties used in the q analysis are larger; this distinction should be stated explicitly.","section":"Section IV"},{"comment":"The bands in Figure A.1 show count ratios for different interaction models, but the text does not specify how the band width is computed (e.g., envelope versus standard deviation across models) or whether the atmospheric-model variation (CORSIKA US Standard versus NRLMSISE-00) is included in those bands.","section":"Appendix A"},{"comment":"The statement that the 7% uncertainty case achieves '50%-73% of the uncertainty in Super-K's sin2θ23 measurement' is unclear; please specify whether this is relative to the 15% benchmark used in the paper or to the published Super-K result, and define the comparison procedure.","section":"Section VI"},{"comment":"The detection efficiency ε is described as 80%, but the paper does not state whether angular resolution and energy resolution are folded into the event-rate calculation or the likelihood; if they are neglected, this should be stated as an approximation.","section":"Equation (1)"}],"recommendation":"major_revision","confidential_remarks":"The core idea is interesting and the closed-loop simulation is competently executed, but the abstract and conclusions go beyond what the test statistic actually computes. Please ensure that the hadronic-interaction uncertainty is either propagated into q and the template fit or explicitly removed from the abstract's claims. The helium-composition assumption also needs quantitative support. These are load-bearing issues but are fixable within the scope of the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the core idea is new and worth engaging—treating a next-generation neutrino detector as a cosmic-ray observatory instead of the other way around. The sensitivity projections are coherent, and the template-unfolding machinery is a sensible way to frame the question. But the abstract says hadronic interaction uncertainties are accounted for, and they are not in the main test statistic; the 20% → 7% flux reduction is also an internal template-fit precision, not a total systematic uncertainty. This deserves refereeing, but the claims need to be aligned with the analysis.\n\nWhat's new: reversing the Evans et al. logic, using MCEq-generated templates to fit Hyper-K's simulated νe and νμ spectra, and extracting discrimination power between AMS, PAMELA, and HKKM models. Full-sky, long-baseline monitoring is a real advantage over balloon and satellite experiments. The energy-dependent shape information is a step beyond a flat rescaling, and the q test against external CR datasets is meaningful. The 5–10% reconstructed flux precision and the factor-of-2 θ23 sensitivity illustration are useful benchmarks for a future experiment.\n\nSoft spots: the biggest is the hadronic-interaction gap. Figure 3's caption says 'holding the interaction model fixed,' and Appendix A shows the SIBYLL/EPOS/QGSJET/DPMJET spread as count-ratio bands, but that spread never enters q. The abstract's claim that the differentiation holds 'even when accounting for uncertainties in neutrino cross sections and hadronic interactions' is not supported by the analysis. At 1–10 GeV, where HKKM New vs AMS is 1.1–1.4σ, a correlated ~5% hadronic shape uncertainty (Ref. [75]) could plausibly erase the signal. This is fixable—either include a hadronic nuisance in q or hedge the abstract.\n\nSecond, the 20% → 7% flux uncertainty reduction is a template-fit precision using templates produced by the same MCEq/HKKM machinery that generates the pseudo-data. It measures how well the seven-parameter template family constrains the injected model, not an absolute calibration of the real flux. The paper does not demonstrate robustness to a spectrum outside that family. Relatedly, the proton-only assumption for the helium component is asserted without propagation; for a 7% claim, a 10–20% helium fraction with a slightly different spectral shape needs a sensitivity check before the number is credible.\n\nThe θ23 study is honestly labeled illustrative, so I don't hold it to the same bar.\n\nWho this is for: people working on atmospheric neutrino backgrounds, oscillation systematics, or future cosmic-ray observables. It's a projection paper—no measurement yet—but it points to a complementary role for kiloton-scale detectors. I'd send it to peer review, because the idea is worth engaging and the flaws are addressable. The right outcome is a revision that brings the abstract in line with the nuisance treatment and adds a helium-robustness test.","headline":"A genuinely reversed CR-to-neutrino sensitivity study with a coherent framework, but the abstract overclaims hadronic-interaction coverage and the headline flux precision is an idealized template-fit number.","tokens_in":13736,"tokens_out":3365,"would_cite":true,"duration_ms":37536,"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":"This paper claims that Hyper-Kamiokande, from atmospheric neutrinos alone, can distinguish competing cosmic ray models, reconstruct the primary spectrum to 5–10% precision, and roughly double sensitivity to the neutrino mixing parameter…","keywords":["atmospheric neutrinos","primary cosmic ray spectrum","Hyper-Kamiokande","spectrum unfolding","neutrino flux uncertainty","hadronic interaction models","neutrino oscillations","theta-23 sensitivity"],"falsifier":"Feed helium primaries at the measured AMS abundance (about 10–20%) through the same MCEq shower calculation instead of treating them as protons, and compare the resulting neutrino spectra between 0.1 and 10 GeV with the proton-only templates; if the helium-inclusive spectra differ by more than the quoted 5–10% per energy bin, the reconstructed primary spectrum in the paper's Figure 5 is biased and the 7% flux-uncertainty claim does not hold.","tokens_in":12749,"feed_emoji":"🌊","tokens_out":7330,"duration_ms":68138,"temperature":0.7,"pith_summary":"This paper tries to establish that next-generation neutrino detectors can work backwards: instead of using cosmic ray measurements to predict the atmospheric neutrino background, use detected atmospheric neutrinos to measure the primary cosmic ray spectrum. The authors show that a 10-year Hyper-Kamiokande exposure would collect enough neutrino events to distinguish competing cosmic ray models (HKKM New at 1.1σ, PAMELA at 2.1σ, HKKM Old at ≥5σ relative to an AMS baseline) despite a 10% cross-section uncertainty. They introduce an unfolding technique that reconstructs the primary spectrum from seven injected proton lines, reducing the neutrino flux uncertainty from about 15–25% to 5–10%, and demonstrate that this improved flux reduces the sin²θ₂₃ uncertainty to 50–73% of Super-Kamiokande's current value. If correct, this turns water Cherenkov detectors into full-sky cosmic ray monitors and removes a dominant systematic from precision neutrino oscillation physics.","feed_headline":"Hyper-K's neutrinos can map the cosmic ray spectrum","feed_subtitle":"Ten years of atmospheric neutrino data would cut flux uncertainty from about 20 percent to 7 percent.","key_machinery":"The load-bearing tool is a template-unfolding scheme: inject single proton lines at seven primary energies (2, 5, 17, 53, 167, 528, and 1670 GeV), propagate each through the MCEq cascade-equation simulation to obtain neutrino flux templates, rescale the HKKM11 and low-energy fluxes, then fit the predicted all-sky Hyper-K event rates with a χ² that carries a 10% Gaussian neutrino-oxygen cross-section systematic. The model-discrimination analysis uses a Poisson maximum-likelihood test statistic q with the overall normalization free and the cross section nuisance-parameterized.","core_discovery":"The paper claims that a kiloton-scale water Cherenkov detector such as Hyper-Kamiokande can, from atmospheric neutrino events alone, act as a cosmic ray observatory. With a 10-year exposure and the neutrino-oxygen cross section fixed to a 10% Gaussian uncertainty, the predicted event rates give statistical power below 1%, enough to tell injected primary models apart: HKKM New at 1.1σ, PAMELA at 2.1σ, and HKKM Old at ≥5σ against an AMS baseline. Fitting seven injected proton lines as neutrino-flux templates unfolds the primary spectrum to 5–10% per energy bin, cutting the atmospheric neutrino flux uncertainty from 15–25% to about 7%. Using that improved flux in a simplified sin²θ₂₃ analysis turns a systematics-limited measurement into one that reaches 50–73% of Super-Kamiokande's current uncertainty, and would approach 1% precision if the flux uncertainty were negligible.","pith_inferences":["If the neutrino-oxygen cross-section uncertainty can be pushed from 10% toward 3%, the same data become sensitive to the helium component of cosmic rays, effectively turning neutrino detectors into composition monitors; that sensitivity is a consequence of the paper's numbers, not a demonstrated result.","A multi-detector joint fit across Hyper-K, JUNO, and DUNE could break the degeneracies the paper notes between neighboring proton-line templates, plausibly pushing reconstruction precision below the 5% floor quoted here.","Because the paper's flux is solar-cycle averaged, a time-resolved version of the unfolding could track solar modulation over the 11-year cycle, using the same machinery with time-binned data."],"forward_implications":["With one 10-year Hyper-K exposure, the atmospheric neutrino sample becomes a continuous full-sky cosmic ray monitor covering primary energies from about 1 GeV to 1 TeV, complementing balloon and satellite missions that sample specific altitudes and times.","Reducing the atmospheric neutrino flux uncertainty from 15–25% to 5–10% removes a dominant systematic for sin²θ₂₃; the paper finds the 7% case reaches 50–73% of Super-K's uncertainty, while a 0% flux uncertainty would push the parameter to about 1%.","The same reduction sharpens background predictions for diffuse supernova neutrino searches, proton decay searches, and dark matter detection, all of which are normalized to the atmospheric neutrino flux.","The analysis is not limited to Hyper-K: the paper states the same reconstruction should work for JUNO and DUNE, whose different locations add independent geomagnetic and solar-modulation systematics."],"supporting_citations":[{"why":"Supplies the AMS proton flux used as the benchmark primary cosmic ray model against which all injected models are compared.","marker":"[36]"},{"why":"Supplies the PAMELA proton flux used as an alternate injection to test model discrimination.","marker":"[38]"},{"why":"Provides the HKKM11 atmospheric neutrino flux calculation whose predictions are rescaled and compared with Super-K data.","marker":"[44]"},{"why":"Provides the low-energy atmospheric neutrino flux model component used in the HKKM comparison.","marker":"[45]"},{"why":"Supplies MCEq, the cascade-equation simulation used to propagate injected primary spectra into neutrino templates.","marker":"[47]"},{"why":"Supplies the site-dependent low-energy flux predictions used below 0.15 GeV where HKKM11 stops.","marker":"[51]"},{"why":"Super-K atmospheric flux measurements that validate the HKKM11 predictions and set the detection-efficiency benchmark.","marker":"[63]"},{"why":"The GENIE neutrino generator tune used for the neutrino-oxygen cross section and its 10% uncertainty.","marker":"[83]"},{"why":"Super-K's quoted flux normalization uncertainties (14.3% below 1 GeV, 7.8% above) that define the current systematic to be improved.","marker":"[90]"},{"why":"Super-K's oscillation analysis with neutron tagging that provides the current sin²θ₂₃ uncertainty benchmark.","marker":"[50]"}],"fun_headline_variants":["Neutrinos turn Hyper-K into a cosmic ray telescope","Hyper-K neutrino data could reveal cosmic ray origins","Atmospheric neutrinos to measure cosmic rays with Hyper-K","Hyper-K's decade of neutrinos sharpens cosmic ray flux","Neutrinos from Hyper-K cut cosmic ray flux error to 7%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The helium component of cosmic rays (10–20% of primaries) is assumed to produce the same neutrino spectrum as protons; if it does not, the reconstructed spectrum and the claimed 7% precision are biased.","fun_headline_variants_meta":{"raw":{"variants":["Neutrinos turn Hyper-K into a cosmic ray telescope","Hyper-K neutrino data could reveal cosmic ray origins","Atmospheric neutrinos to measure cosmic rays with Hyper-K","Hyper-K's decade of neutrinos sharpens cosmic ray flux","Neutrinos from Hyper-K cut cosmic ray flux error to 7%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00023,"raw_usage":{"total_tokens":1463,"prompt_tokens":906,"completion_tokens":557,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":473}},"tokens_in":522,"tokens_out":557,"duration_ms":5608,"temperature":1.0,"reasoning_tokens":473,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:39:38.444048+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Feed helium primaries at the measured AMS abundance (about 10–20%) through the same MCEq shower calculation instead of treating them as protons, and compare the resulting neutrino spectra between 0.1 and 10 GeV with the proton-only templates; if the helium-inclusive spectra differ by more than the quoted 5–10% per energy bin, the reconstructed primary spectrum in the paper's Figure 5 is biased and the 7% flux-uncertainty claim does not hold.","supporting_citations":[{"cited_title":"Time variation of the atmospheric neutrino flux at dark matter detectors","cited_arxiv_id":"2110.14723","evidence_quote":"Supplies the site-dependent low-energy flux predictions used below 0.15 GeV where HKKM11 stops."}],"review_version":1}