{"id":"28e7707f-6c74-47b0-834d-792101eb8440","arxiv_id":"2506.15765","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A sensitivity study shows large-volume neutrino detectors could discover supernova-boosted dark matter down to couplings of about 1e-22, with the Betelgeuse timing signal as a key multi-messenger signature.","lead":"This paper studies dark matter particles that get boosted to high speeds when a supernova explodes, and asks whether DUNE, Hyper-Kamiokande, and JUNO could see them. It projects that these huge neutrino detectors could probe dark matter couplings far smaller than current experiments, especially if a nearby star like Betelgeuse explodes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 90% C.L. sensitivity curves assume zero background, but the paper's own atmospheric-neutrino estimates imply hundreds to thousands of background events in the signal window, so the claimed diffuse-flux reach is not a valid 90% C.L. limit.","rationale":"I identify the background-free assumption, not the low-y extrapolation, as the single most load-bearing concern because the paper's headline reach is quoted from 90% C.L. curves that presuppose zero background. The paper's own background discussion (Section 3.3) estimates 37 atmospheric neutrino events per kt·yr, which for the quoted exposures implies hundreds to thousands of events over 10 years; even with the stated <5% fraction below 100 MeV, the background vastly exceeds the 2.3-event level used for the 90% C.L. contours. The extrapolation of Nχ to small y flagged by the Reader is a real but secondary issue: in the vector-portal model with a heavy mediator, the production rate is proportional to the effective coupling y, so a constant log-log slope of roughly unity is physically expected, and a dedicated calculation would likely confirm the extrapolation. The paper is transparent about the background-free assumption, but the central claim as stated in the abstract and conclusions does not carry the caveat with it. Conditional acceptance remains appropriate, with a quantified atmospheric-neutrino background estimate in the signal window as the required revision.","tokens_in":19632,"tokens_out":17196,"duration_ms":171020,"concrete_test":"Using the Honda atmospheric neutrino flux and standard neutrino-electron elastic and charged-current quasi-elastic cross sections, compute the expected number of electron-like events with recoil energy between the detector threshold and 50 MeV for DUNE-40kt and HK-187kt over a 10-year exposure, including detector efficiency and containment or angular cuts. If the resulting background exceeds ~10 events, the Figures 3 and 6 '90% C.L.' contours should be relabeled as signal-event-rate contours, and the abstract's discovery claim must be amended to state the background dependence explicitly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central projections (Figures 3 and 6, and the abstract's y reach down to ~1e-22) are 90% C.L. contours computed with 2.3 signal events, i.e., an explicitly background-free analysis (Section 3.3). Section 3.3 nonetheless estimates the irreducible atmospheric-neutrino quasi-elastic background at ~37 events per kt·yr before thresholds, with <5% of these events produced by neutrinos below 100 MeV. Taking 1.85 events/kt·yr and the paper's exposures (DUNE 40 kt, HK 187 kt fiducial, 10 years) gives roughly 7.4e2 and 3.5e3 background events, respectively, before any spectral or directional cuts. Even a 10% survival rate leaves the background orders of magnitude above the 2.3-event level used for the 90% C.L. curves. Since the BDM-electron scattering signal is a featureless recoil continuum extending to <50 MeV, similar in shape to neutrino-electron scattering, no cut is demonstrated to reduce this background to ~2.3 events while retaining the signal. Consequently, the claimed '90% C.L. sensitivity' is not established; the 10-, 100-, and 1000-event contours are the defensible projections, which shift the reach to higher y. This concern is more load-bearing than the low-y production extrapolation because it affects all couplings shown, not only the extrapolated tail.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript studies the observability of supernova-boosted dark matter in large-volume neutrino detectors. The authors consider a fermionic dark matter particle coupled to a dark photon, adopt the supernova production yield N_chi(m_chi,y) from the Monte Carlo simulations of Ref. [111], and compute both the diffuse galactic BDM flux and the flux from a single nearby supernova, using Betelgeuse at 197 pc as the example. Detection is via chi-e elastic scattering in DUNE, Hyper-Kamiokande, and JUNO. Event rates are computed with Eq. (4.1), and sensitivity is shown as contours in the (m_chi, y) plane for 2.3, 10, 100, and 1000 signal events over 10 years. A time-delay analysis relating BDM arrival time to m_chi is also presented as a multi-messenger signature. The headline claim is that these detectors could probe effective couplings down to y ~ 1e-22 for dark matter masses around 5-100 MeV.","tokens_in":19944,"tokens_out":20248,"duration_ms":210581,"significance":"The paper addresses a timely and well-motivated gap: previous supernova-boosted dark matter studies focused on ton-scale direct detection via nuclear recoils, whereas the authors emphasize electron scattering in large-volume neutrino detectors. The scattering cross sections and event-rate integrals are standard and internally consistent, and the use of an externally simulated N_chi avoids a self-referential fit of the source. The 10/100/1000-event contours are honest exposure-based quantities, and the multi-messenger time-delay observable is a useful and interesting addition. If the sensitivity claims were fully supported by a background treatment, the projected reach well below current limits would be significant. However, as detailed below, the headline 90% C.L. curves are not established because they ignore the paper's own atmospheric-neutrino background estimate, and the low-y portion of the curves relies on an unvalidated extrapolation of N_chi.","major_comments":[{"comment":"The 90% C.L. sensitivity contours are computed from 2.3 signal events under the explicit assumption of negligible background, but the background estimate given in Section 3.3 rules out that assumption for the diffuse-flux analysis. Taking the quoted 37 atmospheric quasi-elastic events per kt·yr, with less than 5% of those events produced by neutrinos below 100 MeV, gives about 1.85 events per kt·yr; over 10 years this is roughly 740 events for a 40 kt DUNE exposure and 3460 events for the 187 kt HK exposure before any selection cuts. Since the BDM signal is a featureless electron-recoil continuum below about 50 MeV, and no spectral, directional, or other selection is demonstrated to reduce this background to the 2.3-event level while retaining the signal, the curves labeled \"90% C.L.\" (and the y ~ 1e-22 reach quoted in the abstract) are not supported as stated. The 10-, 100-, and 1000-event contours are the defensible projections, or the authors should perform an actual background-inclusive limit calculation.","section":"Section 3.3, Figures 3 and 6"},{"comment":"The low-y part of all sensitivity curves rests on an extrapolation of the production yield N_chi(m_chi,y) to couplings well below the range simulated in Ref. [111], justified only by \"an approximately constant slope on a log-log scale.\" No argument is given that the production rate remains a single power law over the many extrapolated decades in y, and since N_chi enters the flux linearly in Eqs. (2.5), (2.9), and (4.1), a change in the true asymptotic slope would shift the projected y reach by orders of magnitude. The authors should either compute or bound N_chi in this low-coupling regime with a concrete emission model, or restrict the presented reach to the range covered by Ref. [111].","section":"Section 2.1, Figure 1"},{"comment":"The integration domain stated for the diffuse flux is not the galactic volume used to define the supernova rate in Eq. (2.2). With R_max = R_E = 8.7 kpc and z_max = z_E = 24 pc, the integral excludes supernovae with r > R_E and, more importantly, excludes most of the vertical extent of the disk as well as all sources on the opposite side of the galactic plane. For the quoted scale lengths R_d = 2.9 kpc and H = 95 pc, this truncation suppresses the flux normalization by a factor of several, directly affecting the diffuse-flux results in Figures 3 and 4. Please correct the integration limits, or justify the truncation and state the resulting normalization explicitly.","section":"Equation (2.3)"}],"minor_comments":[{"comment":"The text states that DUNE's total mass is 4 x 17.5 kt LAr-equivalent, but the quoted number of target electrons (1.084 x 10^34) and the label \"DUNE-40kt\" in Section 3.3 correspond to a 40 kt fiducial mass. Please clarify which exposure is used and apply it consistently throughout.","section":"Section 3.1 and Table 1"},{"comment":"The sentence \"The number reported in Ref. [111] has typos. We have confirmed this with the authors\" is not appropriate or reproducible in a published paper; please state the corrected normalization explicitly with a clear derivation or reference.","section":"Section 3.3"},{"comment":"The color assignments in the caption (\"DUNE in red, JUNO in red and brown, HK in red, brown, and purple\") make the curves very difficult to distinguish; please use separate colors or directly labeled contours.","section":"Figure 3 caption"},{"comment":"The choice of decoupling temperature T = 30 MeV is stated without justification or discussion of uncertainty, even though the emitted BDM spectrum, the recoil-energy distribution, and the time-delay predictions all depend on T. A brief justification or reference would help.","section":"Section 2.1"}],"recommendation":"major_revision","confidential_remarks":"The paper has two central blockers: the headline 90% C.L. sensitivity curves are background-free while the paper's own background estimates put the diffuse analysis decades away from a background-free regime, and the low-y reach depends on an unvalidated extrapolation of N_chi. There is also a normalization issue in the diffuse flux integration domain that needs clarification. The underlying framework and the 10/100/1000-event contours are useful, so I recommend major revision rather than rejection, asking the authors to either implement a background-inclusive sensitivity treatment or reframe the claims, and to justify or correct the N_chi extrapolation and Eq. (2.3)."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a clean, clearly written sensitivity projection, but the headline reach is not defensible. The authors take the supernova production flux from DeRocco et al., compute BDM-electron scattering at DUNE, Hyper-K, and JUNO, and add a Betelgeuse timing signal. The formalism is standard and the paper is honest about its assumptions. The Betelgeuse time-delay idea is genuinely nice.\n\nThe main problem is the zero-background 90% C.L. contours. The authors themselves estimate roughly 37 atmospheric neutrino events per kt·yr before thresholds, and even after a <5% low-energy cut you get hundreds to thousands of events in a 10-year exposure. A featureless electron recoil continuum is not going to be cut down to 2.3 events without a demonstrated strategy. So Figures 3 and 6 (left panels) are not 90% C.L. sensitivities; they are an idealized signal-counting exercise. The 10, 100, and 1000 event contours are the defensible ones, and they shift the reach to larger y. This is not a minor quibble - the abstract's y down to ~1e-22 is exactly the zero-background projection.\n\nThe other soft spot is the low-coupling extrapolation of the production yield. The constant log-log slope is asserted, not derived, and it directly controls the bottom of the sensitivity curves. The authors disclose this in the caption, but disclosing a guess is not the same as justifying it. A steeper turnover at small y would erase most of the claimed low-coupling region.\n\nWhat is genuinely good: the application to electron scattering at large-volume detectors is a natural extension of Ref. [111]; the timing analysis for Betelgeuse is a concrete multi-messenger idea; and the paper is open about what it did and did not compute. It is not circular - the same coupling enters production and detection, which is standard.\n\nI would send this to peer review, but with the expectation of major revision. The core calculation is probably right, but the presentation needs to demote the 90% C.L. curves to optimistic projections, include a real background estimate, and either justify or remove the low-y extrapolation. As it stands, it is a useful phenomenology note for people working on boosted dark matter, not a discovery claim.","headline":"A clearly written sensitivity projection whose headline 90% C.L. reach is not defensible: the curves assume zero background despite atmospheric neutrino backgrounds orders of magnitude higher, and the low-coupling tail relies on an unvalidated production extrapolation.","tokens_in":20470,"tokens_out":3193,"would_cite":true,"duration_ms":31013,"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":"The paper argues that supernova-boosted dark matter can be detected as MeV-scale electron recoils at DUNE, Hyper-Kamiokande, and JUNO, reaching effective couplings far below current direct-detection limits.","keywords":["supernova-boosted dark matter","boosted dark matter","dark photon","vector portal","DUNE","Hyper-Kamiokande","JUNO","multi-messenger astrophysics"],"falsifier":"Run a dedicated supernova dark-matter production simulation at couplings $10^{-22}$ to $10^{-18}$ and masses $8$--$128$ MeV, and compare the yield with the constant-slope extrapolation from the simulation used in the paper; a drop of more than about an order of magnitude would erase the lowest-$y$ sensitivity contours. A longer-term check: if Betelgeuse explodes and DUNE, HK, or JUNO records no electron-recoil events in the years after the neutrino burst, despite an optimistic prediction of 10 or more events, the central signal claim is ruled out for those parameters.","tokens_in":19439,"feed_emoji":"💥","tokens_out":17651,"duration_ms":154778,"temperature":0.7,"pith_summary":"Core-collapse supernovae can produce and accelerate light fermionic dark matter $\\chi$ through a dark-photon mediator, turning individual stellar explosions into sources of boosted dark matter. The paper projects that the diffuse flux from all galactic supernovae, and the burst from a nearby event such as Betelgeuse, would show up as electron recoils in DUNE, Hyper-Kamiokande, and JUNO. If correct, these detectors would probe effective couplings $y$ between about $10^{-13}$ and $10^{-22}$ for dark-matter masses around $5$ to $100$ MeV, a region well below current direct-detection limits. The paper also identifies a multi-messenger handle: because the dark matter moves slower than light, its signal would arrive months to years after the supernova neutrino burst, tying any excess to the explosion.","feed_headline":"Supernova-boosted dark matter could light up DUNE, HK, and JUNO","feed_subtitle":"A nearby blast's dark-matter pulse would arrive years after the neutrino burst, flagging its origin.","key_machinery":"The central machinery has three parts. The interaction is the vector-portal Lagrangian of Eq. (2.1), in which dark matter $\\chi$ couples to a dark photon through kinetic mixing and a dark gauge coupling; its strength is compactly parametrized by the effective coupling $y = \\epsilon^2 \\alpha_D (m_\\chi/m_V)^4$, which controls production in the supernova, escape from the protoneutron star, and scattering at Earth. The flux is built from the number $N_\\chi(m_\\chi,y)$ of escaping dark-matter particles per supernova, an assumed Fermi-Dirac spectrum at $T = 30$ MeV, and either the galactic supernova spatial distribution or a point source at $197$ pc. The observable is the electron-recoil rate, obtained by convolving the differential $\\chi e^- \\to \\chi e^-$ cross section with the flux, with detector thresholds of $5$ MeV for DUNE and HK and $0.1$ MeV for JUNO. The multi-messenger clock is $\\Delta T = D_S/v_\\chi - D_S/c$, the time delay between the neutrino burst and the arrival of dark matter of mass $m_\\chi$ and energy $E_\\chi$.","core_discovery":"The paper's central claim is that supernova-boosted dark matter is a near-term discovery channel at large-volume neutrino detectors. In a vector-portal model, fermionic dark matter $\\chi$ scatters off electrons through the t-channel exchange of a dark photon $V_\\mu$; the supernova's protoneutron star produces a semi-relativistic $\\chi$ flux with a Fermi-Dirac spectrum at temperature $T = 30$ MeV. Using production yields from the earlier simulation, the paper computes the steady diffuse flux from the galactic supernova population and the flux from a point-like source at the 197 pc distance of Betelgeuse, then folds in the electron-recoil cross sections and detector thresholds of DUNE, Hyper-Kamiokande, and JUNO. The result is a set of 90% C.L. sensitivity contours spanning $m_\\chi \\sim 5$--$100$ MeV and $y \\sim 10^{-13}$--$10^{-22}$, with the single-supernova case reaching smaller couplings than the diffuse case. A Betelgeuse explosion would give a time-delayed dark-matter signal that could be compared with the neutrino burst to measure dark-matter properties.","pith_inferences":["Editorial extension: the region below $y \\sim 10^{-18}$ is only as strong as the constant-slope extrapolation of the supernova production yield; a dedicated simulation at small couplings would either validate or remove that region before any nearby supernova arrives.","Editorial extension: the same time-delay technique applies to other transient astrophysical dark-matter sources, such as neutron-star mergers, where the delay would again encode the dark-matter mass and the source distance.","Editorial extension: a measured dark-matter arrival-time profile after a Betelgeuse supernova would encode $m_\\chi/E_\\chi$, effectively yielding a dark-matter mass measurement rather than just an event excess.","Editorial extension: the diffuse dark-matter signal has no directional tag per event, so verifying it may require comparing electron-recoil energy spectra across DUNE, HK, and JUNO, since each detector has a different target material and threshold."],"forward_implications":["With ten years of exposure, DUNE, Hyper-Kamiokande, and JUNO would register between 10 and 1000 signal events in the optimistic threshold scenario, making neutrino detectors into dark-matter observatories.","A nearby galactic supernova such as Betelgeuse would produce a dark-matter burst delayed by months to years relative to the neutrino burst, providing a multi-messenger confirmation that the events come from the explosion.","The diffuse galactic supernova component gives a steady, all-sky dark-matter flux, so the search does not require a supernova to occur during the experiment's lifetime.","At masses below roughly 20 MeV the projected electron-scattering sensitivity extends to couplings near $10^{-22}$, well below existing direct-detection bounds, so these experiments could discover dark matter even while dedicated direct-detection searches come up empty."],"supporting_citations":[{"why":"It supplies the Monte Carlo values of the dark-matter yield per supernova and the direct-detection sensitivity curves the paper compares against.","marker":"[111]"},{"why":"It provides the double-exponential galactic supernova spatial distribution used in the diffuse flux integral.","marker":"[117]"},{"why":"They give the vector-portal interaction Lagrangian and the dark-matter-electron scattering cross section used for event rates.","marker":"[115, 116]"},{"why":"It provides the atmospheric neutrino background event rates per kiloton-year that the paper uses to estimate backgrounds.","marker":"[34]"},{"why":"It sets the big-bang nucleosynthesis bound that excludes dark-matter masses below 5 MeV.","marker":"[130]"},{"why":"It sets the upper coupling limit below which dark matter is not trapped in the supernova, bounding the search region.","marker":"[133]"},{"why":"They provide the DUNE far detector specifications, thresholds, and angular resolution used in the sensitivity projections.","marker":"[93–97]"},{"why":"They provide the Hyper-Kamiokande detector mass, thresholds, and background rejection parameters.","marker":"[98–100]"},{"why":"They provide the JUNO fiducial mass, energy threshold, and detector characteristics.","marker":"[87–89]"}],"fun_headline_variants":["Supernova-boosted dark matter at DUNE, HK, and JUNO","Dark matter boost from supernovae: signal for neutrino detectors","Supernovae may send dark matter to large-volume detectors","Nearby supernova could reveal dark matter via time delay","Galactic supernova dark matter: a new signal for DUNE, HK, JUNO"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire low-coupling reach assumes that the number of dark-matter particles produced per supernova continues to fall with the same constant log-log slope below the coupling range that was actually simulated, as stated in the caption of Figure 1; if production drops more steeply, the deepest part of the discovery region disappears.","fun_headline_variants_meta":{"raw":{"variants":["Supernova-boosted dark matter at DUNE, HK, and JUNO","Dark matter boost from supernovae: signal for neutrino detectors","Supernovae may send dark matter to large-volume detectors","Nearby supernova could reveal dark matter via time delay","Galactic supernova dark matter: a new signal for DUNE, HK, JUNO"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00024,"raw_usage":{"total_tokens":1540,"prompt_tokens":989,"completion_tokens":551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":455}},"tokens_in":605,"tokens_out":551,"duration_ms":5511,"temperature":1.0,"reasoning_tokens":455,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:33:21.072080+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a dedicated supernova dark-matter production simulation at couplings $10^{-22}$ to $10^{-18}$ and masses $8$--$128$ MeV, and compare the yield with the constant-slope extrapolation from the simulation used in the paper; a drop of more than about an order of magnitude would erase the lowest-$y$ sensitivity contours. A longer-term check: if Betelgeuse explodes and DUNE, HK, or JUNO records no electron-recoil events in the years after the neutrino burst, despite an optimistic prediction of 10 or more events, the central signal claim is ruled out for those parameters.","supporting_citations":[],"review_version":2}