{"id":"cb951bb5-0c10-4a47-a3ca-1a72b0aa1ccb","arxiv_id":"1908.05403","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The adapted IceCube search is most sensitive to the final 10^3 seconds of PBH evaporation, with a local burst rate sensitivity of 3.8 million per cubic parsec per year.","lead":"This IceCube proceedings paper adapts the detector's standard time-dependent point source search to look for neutrino bursts from evaporating primordial black holes. It reports a one-year sensitivity that is far weaker than existing gamma-ray limits but establishes a new neutrino channel for future searches with ten years of data.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Rate-density conversion in Sec. 3 is dimensionally inconsistent: rho_dot = (3/4π) n_sens/d_sens^3 has units pc^-3, not pc^-3 yr^-1, and the abstract's 3.8e6 pc^-3 yr^-1 is therefore not supported.","rationale":"The reader's weakest assumption (three-bin time discretization and Gaussian flare model) is a legitimate and self-acknowledged limitation, but it is an approximation within a proof-of-concept; it plausibly changes the efficiency of the search without invalidating the methodology. The rate-density conversion, by contrast, is an unrecognized dimensional error in the derivation of the paper's headline number. The Sec. 3 formula (3/4π) n_sens/d_sens^3 has no time dimension, and no amount of reinterpretation of the 1-year dataset fixes the missing expected-count factor; the correct Poisson conversion must include T_obs and a threshold such as 2.3. Since the abstract's '3.8×10^6 pc^-3 yr^-1' is the paper's most concrete quantitative claim, the rate-density issue is the most load-bearing defect. I agree with the reader that the paper is a plausible proof-of-concept and should not be rejected outright, so the categorical verdict remains CONDITIONAL (UNCHANGED), but the condition must include a corrected, dimensionally consistent derivation of the burst-rate density.","tokens_in":5268,"tokens_out":9461,"duration_ms":92604,"concrete_test":"Recompute the rate-density curve from the d_sens(tau) values used for Fig. 3a with the standard Poisson conversion rho_dot = N_th / [(4π/3) d_sens^3 T_obs], taking T_obs = 1 yr and N_th = 2.3 (also compare N_th = 1). If the corrected curve does not match Fig. 3b / the quoted 3.8×10^6 pc^-3 yr^-1 — with the expected discrepancy being a factor of order n_sens, the event threshold omitted as a burst count — then the headline numerical claim is not the burst-rate sensitivity and should be restated or removed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central numerical result is the local burst-rate sensitivity, but the equation in Sec. 3 that produces it is not a valid burst-rate conversion. The authors define n_sens as the number of signal events required from one PBH to reach sensitivity, set d_sens = d_ref sqrt(n_ref/n_sens), and then write rho_dot = (3/(4π)) n_sens / d_sens^3. The right-hand side has units pc^-3; no observation time appears, and an event-count threshold is used where an expected number of detectable bursts should appear. For a uniform, Poisson-distributed burst population, the expected number of detectable bursts in exposure T is N = (4π/3) d_sens^3 rho_dot T, so the correct conversion is rho_dot = N_th / [(4π/3) d_sens^3 T], with N_th = 2.3 for a 90% upper limit and T = 1 yr in this analysis. The formula in the paper therefore overstates or otherwise distorts the rate density by a factor related to n_sens and to the missing 1/T. Because the abstract quotes the resulting 3.8×10^6 pc^-3 yr^-1 as the main quantitative achievement, this is more load-bearing than the 3-bin approximation, which the paper itself acknowledges and which affects only the shape of the injected signal.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a proof-of-concept search for neutrino bursts from evaporating primordial black holes (PBHs) with initial masses around 10^15 g, using one year of IceCube track-like events. The Hawking spectrum is time-dependent, so the authors approximate the neutrino fluence over the last 1000 s of the PBH lifetime by three time bins with constant broken power-law spectra. They inject these signals into scrambled background data and use IceCube's unbinned time-dependent point-source likelihood to compute the number of signal events n_sens required for a 90% sensitivity and 5σ discovery potential as a function of burst duration τ. From n_sens they derive a maximum detectable distance d_sens and convert this to a local burst rate density, quoting a sensitivity of 3.8×10^6 pc^-3 yr^-1 at τ ≈ 10^3 s. They compare this with gamma-ray limits and propose that a future analysis with 10 years of IceCube data could improve the sensitivity.","tokens_in":5512,"tokens_out":10688,"duration_ms":97962,"significance":"The paper is significant as the first attempt to search for evaporating PBHs in neutrinos, extending the search to the highest energies and complementing gamma-ray instruments. The adaptation of the standard IceCube time-dependent search is described transparently, and the use of a broken power-law fluence approximation makes the PBH signal model computationally tractable. The qualitative finding that IceCube is most sensitive to the final ~10^3 seconds of evaporation is physically plausible. However, the central numerical result, the local burst rate density, is based on a dimensionally inconsistent conversion in Sec. 3 and is not currently supported; the single-declination treatment and the unquantified three-bin approximation add further uncertainty. If the conversion is corrected and the systematic effects quantified, the method would provide a valuable benchmark for future PBH searches, but the present manuscript needs revision before its headline number can be accepted.","major_comments":[{"comment":"The unnumbered equation after the distance conversion, ρ̇ = 3/(4π) n_sens / d_sens^3, is dimensionally inconsistent with the claimed units of pc^-3 yr^-1: the right-hand side has units pc^-3 (n_sens is dimensionless and d_sens is in pc) and no observation time appears. For a uniform, Poisson-distributed burst population with rate density ρ̇, the expected number of detectable bursts within a sphere of radius d_sens during an exposure T is N = (4π/3) d_sens^3 ρ̇ T, so the correct inversion is ρ̇ = N_th / [(4π/3) d_sens^3 T], with N_th ≈ 2.3 for a 90% confidence upper limit and T = 1 yr. Using n_sens, the number of signal events contributed by a single burst, in place of N_th is not a valid conversion. Because the abstract's headline value 3.8×10^6 pc^-3 yr^-1 is derived from this equation, the central quantitative claim is currently unsupported and must be recalculated.","section":"Sec. 3"},{"comment":"The sensitivity (n_sens, d_sens) is computed for a single declination, δ = 16°, but the paper then quotes a 'local density rate sensitivity' assuming a uniform PBH distribution. For an all-sky rate density, the conversion must be averaged over declination because the effective area and atmospheric neutrino background vary strongly with δ. The manuscript does not describe such a sky-average, so the quoted value 3.8×10^6 pc^-3 yr^-1 is at best a single-declination sensitivity and should not be presented as an all-sky local rate density without further justification.","section":"Secs. 2.2 and 3"},{"comment":"The three-bin approximation of the Hawking time dependence is explicitly acknowledged to be non-optimal, but the paper does not quantify how this discretization biases the computed sensitivity. In addition, the signal is injected using the three-bin time structure while the likelihood fits a Gaussian time profile; the claim that the Gaussian hypothesis is 'effective' is stated without any injection-recovery efficiency or bias test. The manuscript should provide a quantitative test (e.g., the fraction of injected PBH bursts correctly identified, or a comparison of n_sens obtained from the three-bin versus a finer time binning) so that the systematic uncertainty of the central sensitivity number is known.","section":"Secs. 2.1 and 2.2"}],"minor_comments":[{"comment":"The sentence 'The test-statistic defined this way would follow a χ2 distribution where the degrees of freedom reflect the number of fit-parameters' is not correct in general for a likelihood ratio with a parameter at its physical boundary (n_s = 0); since the analysis uses scrambled-data and injection distributions, this theoretical claim should be softened or removed.","section":"Sec. 2.2"},{"comment":"The y-axis label in Fig. 3b gives units of yr^-1 pc^-3, but the rate-density equation in Sec. 3 has units pc^-3; the two should be made consistent.","section":"Fig. 3b and Sec. 3"},{"comment":"The quoted sensitivity 3.8×10^6 pc^-3 yr^-1 lies below the lower limit (10^7 yr^-1 pc^-3) of the y-axis in Fig. 3b; the discrepancy between the quoted number and the plotted curve should be resolved.","section":"Abstract and Discussion vs. Fig. 3b"},{"comment":"The broken power-law approximation for the time-integrated fluence is asserted without a reference or a validation plot; a comparison with a full numerical integration of Eq. (2.1) would support the approximation.","section":"Sec. 2.1"},{"comment":"The citation 'Halzen and Zas [10]' refers to a three-authored paper (Halzen, Keszthelyi, and Zas); the in-text citation should list all authors or use 'et al.'","section":"Sec. 1 and reference list"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is an ICRC proceedings paper, so the appropriate standard is that of a published conference contribution. The rate-density formula error is likely a simple omission of the exposure time and a misuse of n_sens, but it affects the abstract, so the revision must be substantive rather than cosmetic. I recommend asking the authors to recalculate with the correct Poisson conversion and to state the single-declination limitation clearly. No concerns about originality or citation practice."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a reasonable first adaptation of IceCube's time-dependent point source search to neutrino bursts from evaporating primordial black holes, and the authors are appropriately cautious in calling it preliminary. The catch is that the main numerical result, 3.8e6 pc^-3 yr^-1, comes from a rate-density conversion that does not have the right units, so that number is not supported as written.\n\nWhat's new and useful: this is the first IceCube-specific sensitivity estimate for PBH neutrino bursts. The method is sensible — take Hawking fluences in the final ~1000 s from Halzen et al., bin the time-dependent spectrum into three intervals to work around the simulation software's restriction to time-independent spectra, inject into one year of scrambled GFU track events, and run the standard unbinned likelihood. The figures comparing required signal events and distance sensitivity against the standard flare search are informative. The claim that IceCube is most sensitive to the last ~10^3 s is plausible and follows from the combination of a hard spectrum and low background.\n\nThe soft spots are real. The three-bin approximation is explicitly acknowledged and affects the signal shape; that is a limitation, not a fatal one for a proof of concept. More importantly, the conversion in Section 3, rho_dot = (3/4π) nsens / d_sens^3, has units of pc^-3, not pc^-3 yr^-1. There is no exposure time anywhere in it. For a uniform, Poisson-distributed burst population, the expected number of detectable bursts in exposure T is (4π/3) d_sens^3 rho_dot T, so the correct 90%-sensitivity conversion is rho_dot ≈ 2.3 / [(4π/3) d_sens^3 T]. The paper instead inserts nsens, the per-burst event-count threshold, where the Poisson threshold should go, and drops the 1/T. Because the abstract quotes the derived number, this is a load-bearing error. It is easy to fix, but the printed sensitivity should not be quoted until it is redone.\n\nThe comparison with Fermi LAT is honest: the authors note that gamma rays currently constrain the rate density about three orders of magnitude better. That puts the result in proper context.\n\nWho should read this: people working on PBH searches with neutrinos, and anyone building time-domain searches for short bursts. As a proceedings, it is a fair status report. As a standalone result, it needs the unit correction and, ideally, more time bins.\n\nRecommendation: engage with this for the method, not the number. If it is submitted to a journal, send it to review, but expect a major revision that fixes the conversion and perhaps strengthens the signal model.","headline":"A legitimate proof-of-concept for an IceCube PBH-neutrino search, but the headline burst-rate sensitivity is computed from a dimensionally inconsistent formula and needs correction.","tokens_in":6093,"tokens_out":3790,"would_cite":false,"duration_ms":39015,"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":"A neutrino telescope search homes in on the final 1,000 seconds of evaporating primordial black holes, using one year of track-like data to reach a local burst-rate-density sensitivity of…","keywords":["primordial black holes","black hole evaporation","neutrino bursts","time-dependent point source search","unbinned likelihood","track-like neutrino events","TeV neutrino astronomy","dark matter candidates"],"falsifier":"Inject simulated bursts generated with a continuously evolving evaporation spectrum (or with many time bins) into the same one-year scrambled data and compare the 90% sensitivity distance and 5σ discovery distance with the three-bin result; if the distances differ by more than the statistical uncertainty of the pseudo-experiments, the reported $3.8\\times10^{6}\\,\\mathrm{pc}^{-3}\\,\\mathrm{yr}^{-1}$ sensitivity and the claim that the last $10^3$ s dominate would be refuted.","tokens_in":4995,"feed_emoji":"🕳️","tokens_out":8017,"duration_ms":77460,"temperature":0.7,"pith_summary":"Primordial black holes near $10^{15}$ g should be completing their evaporation around the present time, and their final moments should produce a burst of high-energy neutrinos. This proceedings paper adapts a neutrino telescope's standard time-dependent point-source flare search so that it can look for those bursts. It claims that with one year of track-like data the search is most sensitive to the last $10^3$ seconds of a black hole's life, corresponding to emission temperatures around $0.77$ TeV, and reaches a local burst-rate-density sensitivity of $3.8\\times10^{6}\\,\\mathrm{pc}^{-3}\\,\\mathrm{yr}^{-1}$. Because the simulation software can only model time-independent spectra, the intrinsically time-dependent evaporation spectrum is approximated by three time bins; the paper states that this is known to be not optimal. This is the first neutrino-based search for evaporating primordial black holes and can be extended to ten years of data.","feed_headline":"Neutrino search homes in on black holes' final 1,000 seconds","feed_subtitle":"The final thousand seconds of a black hole's life dominate a first neutrino-based search.","key_machinery":"The machinery is the three-bin time-integrated neutrino fluence, built from the instantaneous emission rate of the evaporation spectrum integrated over 0–10 s, 10–100 s, and 100–1000 s to evaporation at a reference distance. Each bin is treated as a time-independent broken power law, with spectral breaks set by the black-hole temperature at the bin boundary; subtracting fluences between times gives the fluence of any interval. This feeds an unbinned time-dependent likelihood with a signal PDF composed of directional, energy, and Gaussian temporal terms, and a test statistic whose 90% threshold defines the sensitivity distance. The three-bin approximation is the load-bearing adaptation, since the detector simulation software accepts only time-independent spectra.","core_discovery":"The paper's central claim is that a standard unbinned time-dependent flare analysis, in which signal events are modeled by a Gaussian time profile, can be repurposed to search for evaporating primordial black holes provided the time-dependent emission spectrum is binned into three constant-spectrum time intervals: 0–10 s, 10–100 s, and 100–1000 s before evaporation. The paper finds that the final $10^3$ s dominate the sensitivity, with the black-hole temperature around $0.77$ TeV, and that one year of track-like events yields a 90% sensitivity distance that converts to a local burst-rate-density sensitivity of $3.8\\times10^{6}\\,\\mathrm{pc}^{-3}\\,\\mathrm{yr}^{-1}$. This is less sensitive than the previously published gamma-ray limits, but it is the first neutrino search at these energies and the paper argues it has a clear path to improvement with ten years of data.","pith_inferences":["My inference: switching from three time bins to a fine time grid would better capture the steeply rising spectrum and could either improve the discovery potential or shift the optimal search window; this is a simulation-driven extension the paper does not carry out.","My inference: the burst-rate-density limit can be translated into an upper bound on the present-day abundance of primordial black holes in the $10^{15}$ g range once a mass-to-rate mapping is assumed; the paper stops at the rate density.","My inference: since neutrinos traverse material that stops gamma rays, applying the same search near the Galactic plane could probe a region where gamma-ray burst searches are blind, provided atmospheric-background modeling stays under control."],"forward_implications":["With the full ten-year sample, the same three-bin analysis is expected to push the local burst-rate-density sensitivity well below the one-year value of $3.8\\times10^{6}\\,\\mathrm{pc}^{-3}\\,\\mathrm{yr}^{-1}$.","The most sensitive window is the final $10^3$ seconds of evaporation, at black-hole temperatures around $0.77$ TeV, which is a different phase of the burst than the windows probed by gamma-ray searches.","A detected burst would yield a neutrino-based measurement of the local rate of evaporating primordial black holes, the first at these energies.","The analysis shows that an existing all-sky flare search can be reused for a new source class with only a spectral-model modification, which keeps a decade-long search computationally tractable."],"supporting_citations":[{"why":"Supplies the black-hole particle-emission formula that generates the time-dependent neutrino fluence.","marker":"[1]"},{"why":"Provides the averaged absorption cross-section used to compute emission and the early estimate of a kilometer-scale neutrino detector's sensitivity to PBH bursts.","marker":"[10]"},{"why":"Sets the gamma-ray local burst-rate-density limit quoted for comparison with the neutrino sensitivity.","marker":"[4]"},{"why":"Defines the unbinned time-dependent likelihood, the Gaussian flare signal model, and the sensitivity/discovery-potential criteria used here.","marker":"[14]"},{"why":"Supplies the one year of track-like events that form the data sample and the scrambled background.","marker":"[11]"},{"why":"Underlie the simulation software that accepts only time-independent spectra, which motivates the three-bin approximation.","marker":"[12][13]"}],"fun_headline_variants":["Neutrino hunt focuses on black holes' last 1,000 seconds","First neutrino search for evaporating primordial black holes","IceCube technique targets black hole evaporation bursts","New neutrino probe for black hole death events","Neutrinos seek black holes' final moments before evaporation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reported sensitivity assumes that the intrinsically time-dependent evaporation spectrum is faithfully represented by three constant-spectrum time bins and a Gaussian flare time profile; the paper itself notes that the three-bin approximation is not optimal.","fun_headline_variants_meta":{"raw":{"variants":["Neutrino hunt focuses on black holes' last 1,000 seconds","First neutrino search for evaporating primordial black holes","IceCube technique targets black hole evaporation bursts","New neutrino probe for black hole death events","Neutrinos seek black holes' final moments before evaporation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000236,"raw_usage":{"total_tokens":1515,"prompt_tokens":965,"completion_tokens":550,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":581,"completion_tokens_details":{"reasoning_tokens":473}},"tokens_in":581,"tokens_out":550,"duration_ms":6021,"temperature":1.0,"reasoning_tokens":473,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:15:11.778514+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Inject simulated bursts generated with a continuously evolving evaporation spectrum (or with many time bins) into the same one-year scrambled data and compare the 90% sensitivity distance and 5σ discovery distance with the three-bin result; if the distances differ by more than the statistical uncertainty of the pseudo-experiments, the reported $3.8\\times10^{6}\\,\\mathrm{pc}^{-3}\\,\\mathrm{yr}^{-1}$ sensitivity and the claim that the last $10^3$ s dominate would be refuted.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the black-hole particle-emission formula that generates the time-dependent neutrino fluence."},{"cited_title":"Halzen, B","cited_arxiv_id":null,"evidence_quote":"Provides the averaged absorption cross-section used to compute emission and the early estimate of a kilometer-scale neutrino detector's sensitivity to PBH bursts."},{"cited_title":"Ackermann et al., Astrophys","cited_arxiv_id":null,"evidence_quote":"Sets the gamma-ray local burst-rate-density limit quoted for comparison with the neutrino sensitivity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the unbinned time-dependent likelihood, the Gaussian flare signal model, and the sensitivity/discovery-potential criteria used here."}],"review_version":1}