{"id":"5d14570d-5849-484a-b667-ac17f0fab7d6","arxiv_id":"2505.09673","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Tau neutrino regeneration lets IceCube constrain dark matter annihilation in Earth's core for masses from 10^5 to 10^10 GeV, setting new upper limits on the spin-independent scattering cross section.","lead":"This paper shows that neutrino telescopes can detect dark matter accumulating in Earth's core even when the dark matter is very heavy, because tau neutrinos regenerate as they cross the Earth. The authors use 7.5 years of IceCube data to set the first upper limits on dark matter scattering for masses up to 10^10 GeV, opening a new mass window for indirect dark matter searches.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Signal event rate uses down-going HESE MC without demonstrating that the downward effective area matches the up-going signal acceptance; this could bias the quoted limits.","rationale":"The reader and I identify the same load-bearing assumption: the signal event rate is computed with the down-going HESE MC response while the background uses the up-going response, and the paper does not demonstrate that these responses are interchangeable. This is not a stylistic or consensus-based objection; it is a concrete correctness risk in the analysis pipeline. The central claim is the derived limit on sigma_SI, and any mismatch in detector response directly rescales the signal expectation, hence the limit. I also note the spectral-index dependence: the signal is dominated by tau neutrinos at energies far above the muon neutrino background, so the effective area must be correct for tau neutrinos specifically, and the paper does not show that the HESE MC flavor response is validated separately for down-going tau neutrinos. A concrete closure test, recomputing the limits with the up-going response for signal, would settle the concern. If the test passes, the paper's central claim likely stands, though the limits could still shift modestly with systematics; this keeps the verdict at CONDITIONAL rather than REJECT, since the issue is a fixable analysis validation rather than a demonstrated logical contradiction.","tokens_in":13937,"tokens_out":1639,"duration_ms":15333,"concrete_test":"Use the public HESE 7.5-year MC release to compute A_eff(E_nu, zenith) for cascade and double-cascade events separately for down-going (zenith less than 90 deg) and up-going (zenith greater than 90 deg) selections, with the same reconstructed-energy bins as the paper. Recompute the signal expectation in Section 4 using the up-going effective area for both signal and background (i.e., propagate the TauRunner surface flux with the HESE up-going response), and compare the resulting 90% CL limit on sigma_SI to the published curve. If the limit shifts by more than the statistical uncertainty quoted in Figures 7 and 10, the directional substitution is not validated and the central claim needs revision.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim is the 90% CL upper limit on sigma_SI derived from a binned likelihood over HESE up-going cascade and double-cascade events. The weakest link is the directional substitution in Section 4: the signal flux is already propagated through Earth by TauRunner, so the authors compute signal events with the down-going HESE MC effective area to avoid double-counting Earth absorption, while background events use the up-going effective area. This is internally coherent only if the HESE down-going effective area and the up-going effective area are identical functions of true neutrino energy and deposited energy for the cascade/double-cascade morphologies. But HESE event selection and reconstruction differ markedly for down-going versus up-going events: the veto logic, the background rejection for atmospheric muons, the angular reconstruction and the cascade energy reconstruction have zenith-dependent efficiencies, and the public MC release does not provide a per-event directional response that can be substituted this way. The manuscript presents no comparison of A_eff for down-going versus up-going events in the relevant energy range (roughly 10 TeV to 1 EeV), nor any closure test showing that the signal acceptance derived from down-going MC reproduces the true up-going acceptance for a propagated flux. If the down-going effective area is, for example, larger because the outer veto suppresses fewer down-going neutrino events, the signal expectation is overestimated and the quoted upper limits on sigma_SI are too strong. The same substitution also affects the IceCube-Gen2 projection, which scales from the same analysis. The concern is concrete and testable, and it sits exactly on the conversion between the propagated flux and the quoted cross-section limits.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes that neutrino telescopes can probe very heavy dark matter (DM) annihilating inside the Earth, provided the DM annihilates to tau neutrinos or tau leptons, because tau regeneration mitigates Earth opacity at energies above ~PeV. Using the public codes chiaroNu and TauRunner, the authors compute the capture rate in the Earth, the annihilation rate, and the propagated neutrino flux for DM masses between 10^5 and 10^10 GeV in the tau+tau- and nu_tau anti-nu_tau channels. They then use 7.5 years of IceCube HESE data, selecting up-going cascade and double-cascade events, and construct a binned Poisson likelihood to set 90% CL upper limits on the annihilation rate and on the spin-independent DM-nucleon cross section. The central quantitative result is a set of upper limits on sigma_SI extending to very heavy DM masses, together with a projection for IceCube-Gen2.","tokens_in":14158,"tokens_out":12188,"duration_ms":141036,"significance":"The qualitative idea is attractive and potentially important: if tau regeneration allows Earth-capture searches to reach DM masses beyond the PeV scale, neutrino telescopes could provide a new probe of ultra-heavy DM. The use of public, community-developed codes (chiaroNu, TauRunner) and public HESE data is a strength, as is the explicit comparison with other Earth-capture and direct-detection limits. The projected Gen2 sensitivity, while based on simplified volume scaling, is a useful outlook. However, the numerical upper limits in Figure 7 rely on a non-standard directional substitution in the event rate calculation and on a background fixed to best-fit values without propagated uncertainties. As presented, the central quantitative claim is not yet fully supported; the paper is a promising proof-of-concept that requires a more careful detector-response treatment.","major_comments":[{"comment":"The signal event rate is computed using the down-going HESE MC selection, while the background uses the up-going effective area. This directional substitution is not justified. The DM signal originates from the Earth's core, i.e. from the nadir (up-going) direction, and the HESE veto and reconstruction efficiencies are direction- and zenith-dependent. No closure test is provided to show that the down-going and up-going effective areas for cascade and double-cascade events are identical over the relevant true-energy and deposited-energy ranges (roughly 10 TeV to 1 EeV). If the down-going acceptance is larger (for example because the outer veto suppresses fewer down-going neutrino events), the signal expectation is overestimated and the resulting upper limits in Figure 7 become artificially strong. The authors should either perform a consistent calculation (e.g., use the up-going MC response with the un-attenuated production flux and let the MC handle Earth propagation, or demonstrate explicitly that the two effective-area choices give identical signal expectations) and estimate the shift in the limits.","section":"Section 4, paragraph beginning 'To calcualte the number of signal events'"},{"comment":"The background is fixed to the IceCube HESE best-fit astrophysical flux (normalization 6.37 and spectral index 2.87) and to the best-fit atmospheric parameters, with no nuisance parameters in the likelihood. The published uncertainties on the spectral index (gamma = 2.87 +0.20/-0.19) and normalization are therefore not propagated into the 90% CL limits on the DM annihilation rate and sigma_SI. Since the DM signal overlaps with the astrophysical neutrino background in energy and flavor, this can underestimate the uncertainties in the quoted upper limits. The authors should either profile or marginalize over the background parameters (with appropriate priors or penalty terms) or demonstrate that the limits are robust against the allowed range of background parameters.","section":"Section 4, Eq. (4.1)"},{"comment":"The test statistic is assumed to follow a chi-squared distribution with one degree of freedom by applying Wilks's theorem. This is problematic because the parameter of interest, Gamma_ann, is non-negative and the null hypothesis Gamma_ann = 0 lies on the boundary of the parameter space, so the standard regularity conditions are not met. In addition, the sample is small (15 up-going cascade events), so the asymptotic approximation may be poor. The 90% CL threshold TS >= 2.71 should be validated with Monte Carlo pseudo-experiments or a boundary-corrected distribution. If the threshold changes, the upper limits in Figures 7 and 10 would shift correspondingly.","section":"Section 4, Eq. (4.4)"}],"minor_comments":[{"comment":"There is a typo: 'To calcualte' should be 'To calculate'.","section":"Section 4, paragraph beginning 'To calcualte the number of signal events'"},{"comment":"The caption repeats a sentence fragment: 'This initial flux serves as the input for near each peak denote the assumed DM particle mass. This initial flux serves as the input for propagation simulations...' The duplicated text should be removed.","section":"Figure 3 caption"},{"comment":"The description of the modification to chiaroNu ('we need to modify current chiaroNu by turning off the decay of tau') is vague. Please provide a more precise account of what was changed and how the modified initial flux was validated.","section":"Section 2, paragraph on chiaroNu"},{"comment":"The statement that the tau regeneration effect is 'emphasized for the first time' in this context should be checked against Ref. [40] and related literature, which also consider neutrinos from Earth-bound DM annihilation and may already include tau regeneration; if not, the novelty claim should be sharpened by explaining the difference.","section":"Section 6"},{"comment":"The extrapolation of the LZ and PandaX-4T limits from m_chi = 10^4 GeV to 10^10 GeV should state the assumptions used (halo model, form factor, and scattering kinematics), since extrapolating direct-detection limits far beyond the experimentally covered mass range can be misleading.","section":"Figure 7"}],"recommendation":"major_revision","confidential_remarks":"The directional substitution in Section 4 is the main correctness risk and should be the focus of the revision. Please ask the authors to provide a closure test or a re-analysis using the up-going MC response with a consistent treatment of Earth propagation. Also verify the claim of being first to include tau regeneration for Earth-capture DM against Ref. [40] and similar works."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a genuinely new result, not a repackaging. The authors take the standard Earth-capture/solar-capture machinery, add tau regeneration through the Earth, and get the first upper limits on spin-independent DM-nucleon scattering for masses 10^5-10^10 GeV from IceCube's public 7.5-year HESE data. The regeneration physics is old—Halzen-Saltzberg and later—but the mass window is new: previous Earth-capture neutrino searches stopped around PeV because of opacity. If the limits are correct, they close a mass window and give other telescopes a target.\n\nWhat's done well: the paper is transparent about its pipeline. chiaroNu for injection spectra, TauRunner for propagation, explicit integration with the public HESE MC, a binned likelihood, and a clear comparison against existing bounds (IceCube, ANTARES, Super-K, Pospelov-Ray). The Gen2 projection is clearly labeled as a background-only volume-scaling estimate with caveats. The authors state what they did and why. That's worth credit.\n\nThe soft spots, in rough order of importance.\n\nFirst, the signal event rate in Section 4 is computed with the down-going HESE effective area, on the argument that the propagated flux should not be re-attenuated by the up-going effective area. That logic is not wrong in principle, but the substitution is only valid if the detector response for cascades and double cascades is zenith-independent. The paper provides no comparison of down-going versus up-going effective areas from the HESE release, no closure test. HESE's veto and reconstruction are direction-dependent; this could plausibly bias the signal expectation by tens of percent, and the direction of the bias is unknown. This needs to be demonstrated, or the limits should be computed with an up-going effective area that has the Earth attenuation removed.\n\nSecond, the background is fixed at IceCube's best-fit astrophysical and atmospheric parameters with no nuisance parameters. That makes the likelihood artificially narrow; including systematics would soften the limits. The authors should say how much this matters.\n\nThird, for the initial flux at these extreme masses they use Pythia EW showering rather than the HDMSpectra treatment that chiaroNu has for lower masses. They mention this but don't estimate the effect on the limit. It could be non-negligible at 10^10 GeV.\n\nThere are minor issues: Section 5 says masses up to 10^11 GeV while the abstract and Figure 7 say 10^10; a typo in \"To calculate\"; and the Gen2 scaling is admittedly crude.\n\nMy overall take: the central idea and the rough numbers are likely correct, but the quoted limits are not yet robust. If the detector-response issue checks out, this is a solid contribution. If it doesn't, the limits could shift.\n\nThis is worth a serious referee. Send it out, with instructions that the directional acceptance be fixed or justified. I'd bring it to a reading group to discuss the veto/effective-area subtlety, but I wouldn't cite the limits in my own work until they're vetted.","headline":"A useful, well-scoped paper that opens a new mass window for Earth-capture DM searches via tau regeneration, but the signal acceptance calculation has a directional-substitution issue that needs a closure test before the limits can be taken at face value.","tokens_in":14779,"tokens_out":8249,"would_cite":false,"duration_ms":86764,"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":"Tau-neutrino regeneration lets TeV–PeV neutrino telescopes see dark matter annihilating in Earth's core up to 10^10 GeV, and 7.5 years of IceCube data already set 90% limits on the cross section.","keywords":["tau neutrino regeneration","very heavy dark matter","Earth core dark matter capture","indirect dark matter search","IceCube high-energy starting events","spin-independent dark matter-nucleon cross section","tau lepton annihilation channels","neutrino telescopes"],"falsifier":"Recompute the signal event rate for the same event sample using the up-going effective areas and detector responses rather than the down-going ones; if the resulting 90% confidence cross-section limits move by more than the Monte Carlo statistical uncertainty, the quoted limits depend on that substitution and would need revision.","tokens_in":13713,"feed_emoji":"🌍","tokens_out":12397,"duration_ms":112725,"temperature":0.7,"pith_summary":"The paper's claim is that tau neutrino regeneration removes the usual ~PeV opacity barrier for indirect dark matter searches in Earth, because a very high energy $\\nu_\\tau$ interacting in the Earth produces a short-lived $\\tau$ that decays back into neutrinos rather than being absorbed. As a result, neutrino telescopes operating at TeV–PeV energies can look for dark matter with masses from 100 TeV to 10 EeV ($10^5$–$10^{10}$ GeV) annihilating in Earth's core into $\\tau^+\\tau^-$ or $\\nu_\\tau\\bar\\nu_\\tau$. Using 7.5 years of IceCube high-energy starting events, the paper reports 90% confidence upper limits on the spin-independent dark matter-nucleon cross section in that mass range, a region where the Earth is opaque to all other neutrino flavors. The takeaway is that the tau channel turns a supposedly blind energy window into an observable one, and existing public data already place new bounds in this regime.","feed_headline":"Tau regeneration opens Earth-core dark matter to 10^10 GeV","feed_subtitle":"IceCube data already put 90% upper limits on these very heavy dark matter masses.","key_machinery":"The mechanism that carries the argument is tau regeneration, implemented through the transport equation $\\frac{d\\vec{\\phi}(E,x)}{dx} = -\\sigma(E)\\vec{\\phi}(E,x) + \\int_E^\\infty d\\tilde E\\, f(\\tilde E,E)\\vec{\\phi}(\\tilde E,x)$, where the redistribution kernel $f$ includes neutral-current down-scattering and the decay of taus produced by charged-current interactions. The initial annihilation spectra are fed into a Monte Carlo propagation code that tracks taus and neutrinos through a layered Earth density profile with a 3 km water-ice layer at the surface. The propagated surface flux is then folded with detector effective areas and event-morphology probabilities to produce expected cascade and double-cascade counts, which enter a binned Poisson likelihood.","core_discovery":"The central discovery is that tau regeneration converts an otherwise unobservable ultra-high-energy signal into a lower-energy, detectable one: $\\nu_\\tau \\to \\tau \\to \\nu_\\tau$ (plus other decay products) repeatedly shifts energy downward as the flux crosses the Earth, so the surface spectrum at TeV–PeV energies retains a signal even when the primary dark matter mass is $10^{10}$ GeV. For both annihilation channels considered, $\\tau^+\\tau^-$ and $\\nu_\\tau\\bar\\nu_\\tau$, the paper obtains 90% confidence upper limits on the spin-independent dark matter-nucleon cross section over $10^5$–$10^{10}$ GeV using the 7.5-year HESE event sample. For masses above $10^7$ GeV the two channel limits converge; below that, the $\\tau^+\\tau^-$ channel gives a somewhat weaker limit because taus lose energy to photonuclear interactions in the Earth's core before they can decay.","pith_inferences":["The same tau-regeneration argument should apply to other high-density dark matter reservoirs, such as the Sun or the Galactic center, where neutrino searches also hit the PeV opacity ceiling; extending this Earth-core analysis to those targets would test how general the effect is.","A direct check of the down-going versus up-going detector-response substitution could be made by recomputing the signal with up-going effective areas; if the resulting limits shift materially, the quoted bounds are analysis-dependent rather than physics-driven.","The near-zero double-cascade background suggests that a targeted search for tau-neutrino double cascades in the full sky might be the most sensitive route at the highest masses; the paper mentions double-cascade analyses as future work but does not perform that search."],"forward_implications":["Earth-core dark matter searches no longer need to stop at about 1 PeV; if annihilation produces taus, masses up to $10^{10}$ GeV are within reach of existing TeV–PeV neutrino telescopes.","Because the limits are set on the annihilation rate and then converted to the spin-independent scattering cross section, the result directly constrains the dark matter-nucleon interaction strength in a mass range far above direct-detection experiments.","For $m_\\chi > 10^7$ GeV, the $\\tau^+\\tau^-$ and $\\nu_\\tau\\bar\\nu_\\tau$ channels give nearly identical limits, so the high-mass bound does not depend strongly on which tau-related channel dominates.","Selecting cascade plus double-cascade events strengthens the upper limit by about 17% compared with using all event morphologies, and a dedicated double-cascade search would add a nearly background-free tau signature.","Scaling instrumented volume to a next-generation detector improves the projected limit by roughly a factor of 3 before including better event reconstruction, so the approach is expected to tighten with future telescopes."],"supporting_citations":[{"why":"introduces tau regeneration as a flux-enhancement mechanism for neutrinos traversing the Earth.","marker":"[41]"},{"why":"propagates tau neutrinos and taus through a layered Earth model with stochastic energy loss, producing the surface fluxes used in the analysis.","marker":"[61]"},{"why":"generates the initial tau-pair and tau-neutrino-pair annihilation spectra that seed the propagation calculation.","marker":"[58]"},{"why":"supplies the 7.5-year high-energy starting event sample and its public detector response used in the likelihood.","marker":"[67]"},{"why":"provides the layered Earth density profile assumed in the propagation simulation.","marker":"[50]"},{"why":"gives the effective volume and annihilation-rate formalism connecting the capture rate to the annihilation rate.","marker":"[48]"},{"why":"defines the spin-independent cross-section scaling and nuclear form factors used for the capture rate.","marker":"[49]"},{"why":"provides the likelihood-ratio test statistic used to set 90% confidence upper limits.","marker":"[72]"}],"fun_headline_variants":["Tau regeneration reveals Earth-core dark matter up to 10^10 GeV","IceCube probes heavy dark matter via tau neutrino regeneration","Tau neutrinos expose very heavy dark matter in Earth's core","Very heavy dark matter unmasked by tau neutrino regeneration"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The signal rate is computed with the detector response for down-going events even though the dark matter neutrinos arrive from below, on the ground that propagation through Earth is already included in the flux; the paper does not show that the two detector responses are interchangeable.","fun_headline_variants_meta":{"raw":{"variants":["Tau regeneration reveals Earth-core dark matter up to 10^10 GeV","IceCube probes heavy dark matter via tau neutrino regeneration","Tau neutrinos expose very heavy dark matter in Earth's core","Very heavy dark matter unmasked by tau neutrino regeneration"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000652,"raw_usage":{"total_tokens":2987,"prompt_tokens":937,"completion_tokens":2050,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":553,"completion_tokens_details":{"reasoning_tokens":1979}},"tokens_in":553,"tokens_out":2050,"duration_ms":17269,"temperature":1.0,"reasoning_tokens":1979,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:27:48.074285+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the signal event rate for the same event sample using the up-going effective areas and detector responses rather than the down-going ones; if the resulting 90% confidence cross-section limits move by more than the Monte Carlo statistical uncertainty, the quoted limits depend on that substitution and would need revision.","supporting_citations":[{"cited_title":"TauRunner: A Public Python Program to Propagate Neutral and Charged Leptons","cited_arxiv_id":"2110.14662","evidence_quote":"propagates tau neutrinos and taus through a layered Earth model with stochastic energy loss, producing the surface fluxes used in the analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the layered Earth density profile assumed in the propagation simulation."},{"cited_title":"Searching for Relic Neutralinos using Neutrino Telescopes","cited_arxiv_id":"hep-ph/9603342","evidence_quote":"gives the effective volume and annihilation-rate formalism connecting the capture rate to the annihilation rate."}],"review_version":1}