{"id":"b8ec53a3-823f-47e7-9338-08f29797d172","arxiv_id":"2411.18431","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"The 1CGH catalogue lists 2,791 Fermi-LAT gamma-ray sources above 10 GeV and identifies 525 where Extragalactic Background Light absorption should be measurable.","lead":"This paper builds a catalogue of 2,791 blazar-like sources detected above 10 GeV by the Fermi satellite over 16 years, and flags 525 of them where background light should absorb a measurable fraction of the highest-energy gamma rays. It also assembles distances for most of these sources, which is what converts absorbed gamma rays into a map of the Universe's transparency.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Redshift uncertainties for the ~62% of sources without spectroscopic redshifts are not propagated into tauEbin, so the headline 525-source absorption subsample is not robust.","rationale":"The reader's verdict is CONDITIONAL and identifies redshift accuracy as the weakest assumption. My stress-test agrees. The central claim has three components: 2791 detections, 62 new detections, and a 525-source absorption subsample. The detection counts are supported by standard likelihood and photon-association criteria, and the expected spurious rate is plausibly small after the 4-photon selection. The 525 subsample, however, depends directly on redshifts whose errors are unquantified. The paper provides z-flags but no error bars, and the tau=0.1 threshold is steep in z. A conservative perturbation test could settle whether the subsample is stable. I do not see an internal inconsistency or an obviously fatal flaw; the concern is a robustness issue, so the conditional verdict stands. Data availability is also a practical concern, but the preliminary GitHub version mitigates it.","tokens_in":27446,"tokens_out":13476,"duration_ms":117487,"concrete_test":"Propagate redshift uncertainties: for every source in the 525-source subsample with z-flag=2, perturb z by ±0.3 (or by the uncertainty quoted in the z-origin reference when provided) and recompute tauEbin with the Saldana-Lopez model; for z-flag=3, recompute at the upper end of the plausible range (e.g., z+0.3) to confirm the flag remains. Count sources whose ABS-flag flips across tau=0.1. Also repeat the whole 525-source selection using the Finke et al. (2010) EBL model. If more than 10% of the subsample flips, or if the Finke-based count differs by more than 20%, the 525 headline is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The 525-source subsample is the central EBL-oriented deliverable (Abstract; Sec. 3.3). It is defined by tauEbin(z, Ebinmax) > 0.1, computed with the Saldana-Lopez model and with redshifts taken from the literature. The paper's own redshift census (Sec. 3.1) shows that only 1062/2791 (38%) have z-flag=1 (spectroscopic); 855 have z-flag=2 (photometric/uncertain) and 210 have z-flag=3 (lower limits). No uncertainty on z is propagated into tauEbin, and the catalogue's z column does not list errors. For z-flag=2 sources, photometric-redshift errors for blazars are typically 0.2–0.5; since tau(z) rises steeply with z, a source near the tau=0.1 boundary can easily flip. The lower-limit z-flag=3 cases are conservative (true z larger only increases tau), but the z-flag=2 population is not. Table B2, the list of strongest absorbers, is populated with many uncertain (?) redshifts (e.g., from Foschini22, Shaw13) without error bars. If a non-negligible fraction of the 525 sources move across the threshold under reasonable redshift perturbations, the headline number is not reproducible. The paper does acknowledge the redshift gap qualitatively but does not quantify its impact on the subsample.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the 1CGH catalogue: 2791 gamma-ray emitters above 10 GeV detected at >3 sigma from 16 years of Fermi-LAT data, selected from multifrequency seed positions (5BZcat, 3HSP, TeVcat, 4LAC-DR3, and unassociated 4FGL sources). A binned likelihood analysis in the 10-800 GeV band is followed by a photon-association cut requiring at least four source-type events within 0.12 degrees. The catalogue includes 62 new detections, an extensive literature-based redshift review with quality flags, and an absorption subsample of 525 sources where tau(Ebinmax,z) > 0.1 according to the Saldana-Lopez EBL model. The paper also introduces Ebinmax, the mean energy of the four highest-energy photons, as a robust estimator of the highest detectable energy bin.","tokens_in":27755,"tokens_out":5243,"duration_ms":44967,"significance":"If the catalogue is robust, it is a valuable community resource: it extends >10 GeV detections to fainter fluxes than 3FHL, provides a systematic redshift-quality flagging system, and supplies a well-defined target list for EBL studies and optical follow-up. Strengths include the clear description of the reduction pipeline, the use of public Fermi-LAT data and public catalogues, the explicit estimate of the pre-cut spurious rate, and the extensive literature compilation. However, the central EBL-oriented deliverable, the 525-source absorption subsample, depends on two fragile ingredients: redshift values that lack propagated uncertainties for ~38% of the sample, and the choice of a single EBL model. The paper itself acknowledges the redshift incompleteness qualitatively but does not quantify its impact on the headline number; the post-cut false-positive rate is likewise not quantified. These issues affect the reproducibility of the main claims, though they are addressable within the manuscript's scope.","major_comments":[{"comment":"The ABS-flag and tauEbin values are computed from literature redshifts without propagating redshift uncertainties. The redshift census in Section 3.1 shows that 855/2791 sources (30.6%) have z-flag=2 (photometric/uncertain) and 210/2791 (7.5%) have z-flag=3 (lower limits), and the catalogue's 'z' column lists no errors. Because tau(E,z) rises steeply with z, a source near the tau=0.1 boundary can flip between ABS-flag=1 and 0 under a plausible photometric redshift error of 0.2-0.5. The paper acknowledges the redshift gap qualitatively but does not quantify its impact on the 525-source subsample. Please add a robustness analysis: for example, recompute tauEbin with z +/- sigma_z using typical photometric uncertainties and report how many sources change ABS-flag, or provide a Monte Carlo. For Table B2, mark sources whose ABS-flag determination is fragile.","section":"Section 3.3, Table B2"},{"comment":"The pre-selection spurious rate for TS>12 is estimated as 20.5/3004 ~ 0.68%, but the paper does not quantify the false-positive rate after the additional requirement of at least four associated photons within 0.12 deg (Section 2.4). The statement that the rate 'should be lower' is not a measurement; the photon cut is an additional selection layer applied to the same data and the matching radius is small, so a background-only Monte Carlo or a source-scrambling test is needed to estimate the post-cut contamination. Without this, the reliability of the 62 new detections and of the full 2791-source catalogue is not fully established.","section":"Sections 2.3-2.4"},{"comment":"The ABS-flag selection is based on the Saldana-Lopez et al. (2021) EBL model only, yet the text claims that the resulting subsample is where 'the EBL optical depth can be robustly measured' and that tau values can be derived 'independently of any specific EBL model.' Figure 4 compares the tau=1 curves of four models, but the tau=0.1 boundary used for the 525-source subsample is shown only for one model. Please quantify the model dependence: report how many of the 525 sources remain in the tau>0.1 regime under the Finke et al. (2010), Dominguez et al. (2011), and Franceschini & Rodighiero (2017) models. If the overlap is small, the abstract's 525 number should be presented with a caveat.","section":"Section 3.3"}],"minor_comments":[{"comment":"The captions state 'approximately one-third of the 1CGH sources lack assigned redshift,' but Section 3.1 reports 664/2791 = 23.7% without redshift; the captions are inconsistent with the text and should be corrected.","section":"Figure 4 and Figure 5 captions"},{"comment":"The caption says 'The first three columns show ...', but the table has more than three columns; also the symbols '?' and '>' in the 'z' column are not defined in the caption, although they are explained later in Section 3.1.","section":"Table 1 caption"},{"comment":"The paper states that a complete table is available in the online version and a preliminary version on GitHub; for reproducibility, a machine-readable full catalogue should be included as supplementary material with the submission, including redshift uncertainties.","section":"Data Availability"},{"comment":"The 26 sources whose photon index reaches the Gamma=6.5 limit and the 27 sources with 3FHL curvature are listed with fluxes that should be interpreted as upper limits; the catalogue columns should carry a flag that propagates to the tauEbin analysis if any of these sources are in the absorption subsample, so users do not treat the power-law flux as a detection.","section":"Section 3 (26 and 27 sources)"},{"comment":"The TS-to-sigma conversion is cited to a GitHub repository; for a journal publication, a formal reference or equation should be included in the text.","section":"Section 2.3"},{"comment":"The new Ebinmax estimator (mean of the four highest-energy photons) is introduced without a discussion of its statistical properties or a comparison against the highest-energy photon; a brief justification or a reference to a validation test would help.","section":"Section 3.3 (Ebinmax definition)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a catalogue paper with strong practical value, but the redshift-uncertainty sensitivity of the 525-source subsample and the unquantified post-cut spurious rate need to be addressed before acceptance. The stress-test concern about redshift propagation largely lands; the issues are fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good catalogue paper, worth taking seriously. The 1CGH compilation does what it says: 2791 sources above 10 GeV from 16 years of Fermi-LAT, 62 genuinely new detections, and a careful redshift review that assigns quality flags across nearly 70 literature sources. The redshift census alone is a useful community resource.\n\nWhat is actually new: the catalogue itself, the z-flag system, and the Ebinmax statistic (mean of the four highest-energy photons rather than just the maximum). The pipeline is standard seed-based binned likelihood, but that is fine for a catalogue. The 3FHL comparison shows they reach fainter fluxes, and the paper is transparent about the 53 sources (26 at the spectral-index limit, 27 with 3FHL curvature) whose fluxes should be read as upper limits.\n\nSoft spots: the 525-source absorption subsample is the one place where the analysis could mislead. tauEbin is computed at the literature redshift, and for the ~62% of sources without spectroscopic redshifts - 855 photometric/uncertain, 210 lower limits - no uncertainty is propagated. A photometric redshift error of 0.2-0.5 can push a source across the tau=0.1 boundary, so the ABS-flag list will have non-negligible churn. The authors acknowledge the redshift gap qualitatively but do not quantify its effect on the subsample. That is the weakest link, and it is a moderate weakness, not fatal to the catalogue. Also minor: the spurious-rate estimate covers only the pre-photon-cut stage; the post-cut rate is asserted but not measured. And the catalogue is not yet public, which limits immediate reproducibility. I also spotted a clear typo in Table B1: 3HSPJ213004.8-563222 has an AllWISE ID matching J182756.54+263313.2, obviously wrong.\n\nOverall, the central claim - that these are real >10 GeV detections - holds up. The absorption subsample should be treated as indicative, not final. This paper is for anyone working on EBL measurements, CTAO target lists, or blazar redshift follow-up. It deserves a serious referee; the needed fixes are quantitative robustness checks, not a rewrite.","headline":"A useful and honest catalogue with one soft spot: the absorption subsample inherits unquantified redshift uncertainties.","tokens_in":28304,"tokens_out":2181,"would_cite":true,"duration_ms":20532,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The 1CGH catalogue lists 2791 blazars and candidates detected above 10 GeV at greater than 3 sigma significance from 16 years of Fermi-LAT observations, including 62 new detections and a 525-source subsample where the extragalactic…","keywords":["gamma-ray astronomy","Fermi-LAT","blazars","extragalactic background light","cosmic gamma-ray horizon","1CGH catalogue","high-energy gamma-rays","redshift surveys"],"falsifier":"Stack the highest-energy bins of the 525 ABS-flag sources and compare the average observed flux with the Saldana-Lopez EBL prediction: if the expected suppression exp(-tau) is absent at the predicted energies, the absorption classification is wrong.","tokens_in":27243,"feed_emoji":"🔭","tokens_out":6326,"duration_ms":49263,"temperature":0.7,"pith_summary":"The paper sets out to build the most complete targeted catalogue of gamma-ray emitters above 10 GeV from 16 years of Fermi-LAT data, and to use it to map the cosmic gamma-ray horizon. It reports 2791 detections at better than 3 sigma, 62 of them new, after seeding a likelihood analysis with blazar positions from major catalogues. For every source it reviews redshift information from the literature and assigns a quality flag, then uses the mean energy of the four highest-energy photons to define the highest detectable energy bin. Applying the Saldana-Lopez EBL model, it identifies 525 sources where the optical depth exceeds 0.1, meaning more than 10 percent of the flux at that bin should be absorbed. If this holds, the catalogue gives observers a concrete target list for measuring extragalactic background light, testing redshift completeness, and probing the transparency of the Universe to very high energy photons.","feed_headline":"A 2791-source catalogue maps the cosmic gamma-ray horizon","feed_subtitle":"Sixteen years of Fermi-LAT data reveal 62 new detections and 525 blazars whose gamma-rays are measurably absorbed.","key_machinery":"The load-bearing machinery is the binned likelihood analysis in the 10-800 GeV band with a power-law model, combined with a photon-association step that requires at least four source-type events within 0.12 degrees. The central derived quantity is Ebinmax, the mean energy of the four highest-energy photons, which provides a robust estimate of the highest energy bin a source has; comparing this to the Saldana-Lopez EBL model through the relation F_observed = F_intrinsic exp(-tau(E,z)) yields the tauEbin and ABS-flag columns. The z-flag system, which separates spectroscopic, photometric or uncertain, and lower-limit redshifts, is what makes the absorption classification possible.","core_discovery":"The central discovery claimed is that a large, carefully curated sample of blazars and blazar candidates can be detected above 10 GeV and ranked by expected EBL absorption. The 1CGH catalogue contains 2791 sources detected at >3 sigma over 10-800 GeV, with 62 previously unreported gamma-ray detections, most of them high-synchrotron-peak blazars from the 3HSP catalogue. After an extensive redshift review covering nearly 70 publications, 38.2 percent of the sources have spectroscopic redshifts, 30.6 percent have photometric or uncertain values, 7.5 percent have lower limits, and 23.7 percent have none. Using the Saldana-Lopez EBL model and the mean energy of the four highest-energy photons as Ebinmax, the paper identifies 525 sources with tau(Ebinmax, z) > 0.1, flagging them as cases where moderate to severe absorption should be measurable with Fermi-LAT.","pith_inferences":["The ABS-flag subsample should be treated as provisional: roughly 62 percent of the catalogue lacks robust spectroscopic redshifts, so the 525-source list would change if lower limits or photometric redshifts are revised.","A direct test of the catalogue's utility is to stack the 10-800 GeV spectra of the 525 flagged sources and check that the observed attenuation matches the Saldana-Lopez prediction; a mismatch would point to either intrinsic spectral features or the need for revised EBL models.","Because the selection is seeded by known blazars, the 1CGH is biased against gamma-ray emitters without identified counterparts; unassociated 4FGL sources are included, but a fully unbiased horizon map would require an independent all-sky search.","The Ebinmax statistic could be applied to other energy bands or future missions to extend the horizon mapping beyond Fermi-LAT's reach."],"forward_implications":["The 525 ABS-flag sources form a ready-made target list for measuring the EBL optical depth with Fermi-LAT spectra across redshifts 0 to 3.","The 62 new detections, mostly high-synchrotron-peak blazars, are promising targets for the Cherenkov Telescope Array Observatory and follow-up optical campaigns.","The redshift review, with quality flags, improves the reliability of any EBL or opacity study built on 3FHL and 4LAC-DR3 data.","Sources flagged with z-flag=0 but with optical or radio associations are prioritized for spectroscopy; if redshifts are obtained, they can be added to the measurable-absorption sample.","The catalogue extends the significance distribution toward fainter sources, lowering the effective detection threshold above 10 GeV relative to 3FHL."],"supporting_citations":[{"why":"Supplies the 3FHL high-energy catalogue whose seeds, acceptance criteria, and detection statistics anchor the 1CGH design.","marker":"Ajello et al. 2017"},{"why":"Provides the 5BZcat blazar catalogue, one of the primary seed-position sources for the targeted analysis.","marker":"Massaro et al. 2015a"},{"why":"Provides the 3HSP high-synchrotron-peak blazar catalogue, the source of most of the 62 new gamma-ray detections.","marker":"Chang et al. 2019"},{"why":"The 4FGL-DR4 point-source catalogue supplies associations, background models, and the unassociated high-latitude seed positions.","marker":"Ballet et al. 2023"},{"why":"The 4LAC-DR3 AGN catalogue contributes blazar and blazar-candidate seeds along with photometric redshift estimates.","marker":"Ajello et al. 2022"},{"why":"The reference EBL model used to compute tauEbin and define the ABS-flag threshold of tau > 0.1.","marker":"Saldana-Lopez et al. 2021"},{"why":"Defines the Test Statistic and the point-source detection logic used for candidate selection.","marker":"Mattox et al. 1996"},{"why":"The TeVcat list of TeV-detected sources adds additional seed positions to the candidate sample.","marker":"Wakely & Horan 2008"}],"fun_headline_variants":["2791 blazars map the cosmic gamma-ray horizon","16 years of Fermi data yield 2791 gamma-ray sources","New catalogue: 525 blazars show gamma-ray absorption","Mapping the gamma-ray horizon with 2791 Fermi-LAT detections","62 new gamma-ray sources found in 16-year Fermi dataset"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The absorption classification assumes the literature redshifts are accurate enough for the many sources that lack spectroscopic determinations; if those redshifts are wrong, the computed optical depths and the 525-source subsample change.","fun_headline_variants_meta":{"raw":{"variants":["2791 blazars map the cosmic gamma-ray horizon","16 years of Fermi data yield 2791 gamma-ray sources","New catalogue: 525 blazars show gamma-ray absorption","Mapping the gamma-ray horizon with 2791 Fermi-LAT detections","62 new gamma-ray sources found in 16-year Fermi dataset"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000588,"raw_usage":{"total_tokens":2822,"prompt_tokens":1067,"completion_tokens":1755,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":683,"completion_tokens_details":{"reasoning_tokens":1668}},"tokens_in":683,"tokens_out":1755,"duration_ms":11891,"temperature":1.0,"reasoning_tokens":1668,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:13:01.363720+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Stack the highest-energy bins of the 525 ABS-flag sources and compare the average observed flux with the Saldana-Lopez EBL prediction: if the expected suppression exp(-tau) is absent at the predicted energies, the absorption classification is wrong.","supporting_citations":[],"review_version":1}