{"id":"e176cc41-e3ac-4258-9eca-f8d069b58218","arxiv_id":"1908.06831","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A three-year HAWC survey of 138 nearby active galaxies detects only Mrk 421 and Mrk 501 and sets upper limits on the others, with M87's average TeV flux below Fermi-based expectations.","lead":"Using three years of data from the HAWC gamma-ray observatory, this paper looks for TeV light from 138 nearby active galaxies and clearly detects only the two closest blazars, Mrk 421 and Mrk 501. It then places upper limits on the rest and reports that the radio galaxy M87 is fainter in its average TeV state than Fermi extrapolations suggest.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Upper limits depend on fixed spectral index α=2.5; the M87 comparison could shift if recomputed with M87's actual index.","rationale":"The reader identified the fixed α=2.5 and the EBL approximation as the weakest assumption; I agree with the α=2.5 part but not the EBL part for the M87 claim. The paper's own Mrk 501 numbers show that integrated fluxes derived with α=2.5 can differ by ~70% from those using the best-fit index, so the same effect could shift the non-detection upper limits and the M87 comparison. The M87 result is the scientifically most interesting claim, and it is not affected by EBL, so the test should target the spectral index. The detections of Mrk 421 and Mrk 501 are highly significant and robust. The analysis is standard HAWC maximum-likelihood, but the lack of systematic uncertainties and the simplified EBL approximation for high-z sources are additional caveats. A recomputation using per-source 3FHL spectral indices is a straightforward check that would settle the concern.","tokens_in":6532,"tokens_out":14868,"duration_ms":142614,"concrete_test":"Recompute the M87 95% CL upper limit on N(>0.5 TeV) using the 3FHL best-fit spectral index for M87 (and its 1σ range) in the same HAWC likelihood instead of the fixed α=2.5. If the recomputed limit exceeds the 3FHL extrapolation in the 0.5–2.0 TeV band, the M87 comparison fails. Repeat for all Table 1 non-detections using their 3FHL indices to quantify the systematic spread of the reported limits.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption is the fixed spectral index α=2.5 used for all non-detection upper limits in Table 1 (Section 5.1, Eq. 5.1). The paper's own fits demonstrate the sensitivity: for Mrk 501 the best-fit index is α=1.96±0.10 with φ1=6.8, while the α=2.5 row gives φ1=12.7, changing the integrated N(>0.5 TeV) by roughly 70%. For the non-detected sources the true spectra are unknown, and the 3FHL catalog provides per-source indices that often differ from 2.5. The highlighted claim that the HAWC upper limit for M87 lies below the 3FHL extrapolation is therefore not robust until that limit is recomputed with M87's actual spectral index. The EBL exponential approximation (Eq. 3.1) is a secondary concern and is negligible for M87 at z=0.004.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a survey of active galactic nuclei conducted with the HAWC gamma-ray observatory using 1017 days of data. From a sample of 138 Fermi-LAT 3FHL AGN with redshift z≤0.3, the analysis detects the two nearest BL Lacs, Mrk 421 and Mrk 501, at high significance, and reports 95% confidence-level upper limits for the ten best non-detected candidates under the assumption of a fixed intrinsic spectral index α=2.5. The paper highlights that the HAWC limit on M87 lies below the 3FHL spectral extrapolation and below historical high/intermediate VHE states, and makes analogous statements for IC 310.","tokens_in":6723,"tokens_out":11013,"duration_ms":106877,"significance":"If the results are robust, the survey provides a valuable long-term, time-averaged census of TeV emission from nearby AGN, complementing pointed observations with Cherenkov telescopes. The two detections are extremely significant and consistent with known sources, and the upper limits reach integrated fluxes of order 10^-12 cm^-2 s^-1 above 0.5 TeV. The M87 and IC 310 limits are of particular interest for constraining the quiescent emission of radio galaxies. The analysis uses standard maximum-likelihood methods and the Feldman-Cousins prescription, which are appropriate for this type of search. However, the quantitative cross-comparisons between HAWC limits and Fermi/ACT spectral extrapolations are conditional on the assumed spectral shape, and the paper does not currently demonstrate that key conclusions are robust to that assumption.","major_comments":[{"comment":"The upper limits are computed with the spectral index fixed to α=2.5. The paper's own fits show that this choice has a substantial effect on the integrated flux: for Mrk 501, fitting α=1.96 yields φ1=(6.8±1.1)x10^-12 TeV^-1 cm^-2 s^-1, whereas the α=2.5 fit gives φ1=12.7x10^-12 TeV^-1 cm^-2 s^-1, changing N_obs(>0.5 TeV) by roughly a factor of 1.7. For non-detected sources the true indices are unknown, and the 3FHL catalog provides per-source spectral indices that often differ from 2.5. The highlighted claim in Section 5.2 that the HAWC upper limit for M87 lies more than half below the 3FHL extrapolation is therefore not demonstrated until the M87 limit is recomputed with M87's actual 3FHL spectral index or a range of plausible indices. Please provide this recomputation, state the resulting factor relative to the 3FHL extrapolation, and assess whether the conclusion survives.","section":"Section 5.1, Eq. (5.1); Table 1; Section 5.2"},{"comment":"The reported N_obs(>0.5 TeV) limits also depend on the exponential EBL approximation with z_h=0.108, which is calibrated for α=2.5 and E0=0.5 TeV. The paper notes that z_h varies from 0.095 (α=2.0) to 0.118 (α=3.0), which propagates into a small but not entirely negligible uncertainty in the quoted limits. The paper should state explicitly that all integrated-flux limits in Table 1 are defined for α=2.5 only, and ideally quote a systematic range from the α-dependence of the EBL attenuation and the integral conversion.","section":"Section 5.1, Eq. (3.1); Table 1"}],"minor_comments":[{"comment":"Several significances differ between the text and Table 1: the text quotes s=+45.6 for Mrk 421 and s=+20.2 for Mrk 501, while Table 1 lists +45.8 and +19.9. Please make these values consistent.","section":"Section 5.2 and Table 1"},{"comment":"The abstract states that the analysis 'sets limits for the rest of the sample', but the paper presents limits only for the ten best non-detected candidates (Table 1). Please clarify whether upper limits were computed for all 138 sources and either show a summary of those limits or revise the wording to specify that limits are reported for the selected best candidates.","section":"Abstract and Section 5.2"},{"comment":"The phrase 'with a & 20kHz rate' appears to be a typesetting error; it should read 'with a ≥ 20 kHz rate'.","section":"Section 2"},{"comment":"In the M87 discussion, the phrase 'more than half below the LAT extrapolation' would be clearer as 'more than a factor of two below the 3FHL spectral extrapolation', and the paper should specify which spectral model and energy band are used for the comparison.","section":"Section 5.2"},{"comment":"Reference [13] lists the first author as 'Abeysekara, U.' but the standard author name is 'Abeysekara, A.U.'; please correct this in the bibliography.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is an ICRC proceedings contribution, so the space constraints are understood. The central detections and the upper-limit methodology are sound, but the paper's most interesting quantitative claims about M87 and IC 310 rest on the fixed α=2.5 assumption. A recomputation with the actual spectral indices would settle whether those claims survive; I believe this is achievable within the manuscript's scope and should be required before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe quick take: this is a competent survey paper, not a breakthrough. HAWC confirms TeV emission from Mrk 421 and Mrk 501 and sets 95% upper limits for ten nearby AGN that were not detected. The genuinely new content is the limit set, in particular the M87 and IC310 limits, with observed fluxes around 10^-12 cm^-2 s^-1. Those limits are below the 3FHL spectral extrapolation and below the historical flaring states reported by ACTs. That is a useful addition because most TeV AGN measurements come from pointed observations during activity, and a time-averaged limit from a wide-field monitor is a different kind of information.\n\nThe analysis is standard HAWC maximum-likelihood point-source fitting with Feldman-Cousins limits. The detections are not subtle: s = +45.8 and +19.9. The paper does not oversell them; it treats them as validations. The EBL treatment is deliberately crude—a one-parameter exponential with z_h = 0.108, taken from Dominguez 2011—adequate for a z<=0.3 sample, and the paper states that the redshift cutoff is set by this approximation. I do not see a circularity problem; the attenuation model is external and the detections are known sources.\n\nThe main caveat is the fixed alpha=2.5 used for every upper limit. The paper itself shows the sensitivity: for Mrk 501, alpha=2.5 raises the normalization by about 70% compared with the best-fit index. For non-detected sources, the true indices are unknown, and several 3FHL spectra are hard. In particular, the highlighted M87 limit should be recomputed with M87's 3FHL spectral index (or a bracketing range) before claiming robustly that it lies below the LAT extrapolation. That is an addressable issue, not a fatal one—the limit is what it is under a stated assumption, and the qualitative conclusion that HAWC sees little or no quiescent TeV emission from these AGN stands.\n\nOther soft spots are minor: statistical errors only, no public analysis artifacts, and a proceedings-level description of the candidate selection. None of this changes the central result.\n\nI would bring this to reading group and would cite the M87/IC310 limits. For peer review, send it out; it deserves a serious referee, with a request to add a spectral-index sensitivity check for the upper limits.","headline":"Useful survey result rather than a breakthrough: HAWC confirms the two known TeV blazars and sets a set of new TeV upper limits, with the M87 and IC310 limits being the most valuable, provided you accept the fixed spectral-index caveat.","tokens_in":7313,"tokens_out":2614,"would_cite":true,"duration_ms":28862,"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":"Three years of HAWC data detect only Mrk 421 and Mrk 501 among 138 nearby AGN, and push M87 below the Fermi extrapolation.","keywords":["active galactic nuclei","very-high-energy gamma-ray astronomy","HAWC Observatory","BL Lac objects","M87","Mrk 421","Mrk 501","extragalactic background light"],"falsifier":"Reanalyze the same 1017 days of HAWC data at the position of M87 with simulated source injections at the 3FHL-extrapolated flux: if the likelihood analysis cannot recover such a source at the claimed sensitivity, the upper limit is not a valid test of that extrapolation. Alternatively, a future wide-field TeV exposure with longer integration that measures M87's time-averaged N(>0.5 TeV) above 0.95 x $10^{-12}$ $cm^{-2}$ $s^{-1}$ would directly contradict the reported limit, assuming the source's baseline state is unchanged.","tokens_in":6350,"feed_emoji":"🔭","tokens_out":12644,"duration_ms":122707,"temperature":0.7,"pith_summary":"Almost all very-high-energy observations of active galactic nuclei come from short, pointed runs with Cherenkov telescopes, which can catch flares but say little about what the sources do on average. This paper analyzes 1017 days of continuous HAWC exposure of 138 AGN selected from the 3FHL catalog with redshift z <= 0.3, and establishes the first systematic wide-field picture of their persistent TeV emission. The result: only the two nearest BL Lacs, Mrk 421 and Mrk 501, are detected at high significance, while all 136 other AGN are consistent with no steady emission, with 95% confidence upper limits reaching N(>0.5 TeV) around $10^{-12}$ $cm^{-2}$ $s^{-1}$. The most pointed claim concerns the radio galaxy M87, whose HAWC limit lies below the spectral extrapolation of the 3FHL data and below the 2005-2010 pointed-telescope measurements, implying that M87's quiet TeV baseline is fainter than flaring-state observations suggest. If this holds, it means the TeV sky's extragalactic population is dominated by a handful of nearby objects, and that the variability seen by pointed telescopes is not representative of steady output.","feed_headline":"HAWC confirms two blazars and sets deep TeV limits on 136 AGN","feed_subtitle":"Three years of HAWC data fix M87's baseline below the Fermi 3FHL extrapolation and push 136 AGN below 10^-11 photons per cm^2 s.","key_machinery":"The load-bearing instrument is HAWC's wide-field water-Cherenkov array, whose 300 tanks survey two-thirds of the sky every sidereal day with a median energy response around a few TeV; events are split into nine fhit bins (fhit being the fraction of channels recording a signal, used as an energy proxy), with gamma-hadron separation applied per bin. The analysis uses a standard maximum-likelihood fit of an intrinsic power-law spectrum, (dN/dE)_intr = phi_1 (E/1 TeV)^{-$\\alpha$}, with EBL attenuation $e^{{-tau(E,z)}}$ folded in. For upper limits, $\\alpha$ is fixed to 2.5 and the normalization phi_1 is fit, with 95% confidence intervals from a Feldman-Cousins prescription; the EBL attenuation is reduced to the exponential approximation N_obs(>0.5 TeV) = $e^{{-z/0.108}}$ N_intr using the EBL model adopted by the paper. This machinery converts HAWC counts into astrophysical flux limits that can be placed on the same plot as 3FHL spectra and pointed-telescope measurements.","core_discovery":"The authors report that after 1017 effective days, HAWC detects time-averaged TeV emission from Mrk 421 (significance +45.8) and Mrk 501 (significance +19.9), the only two AGN in the survey with $\\sqrt$(TS) >= 5 in all five 3FHL energy bands. For every other source, the fit returns a 95% confidence upper limit on the observed flux above 0.5 TeV, computed for an assumed intrinsic power law dN/dE = phi_1 (E/1 TeV)^{-$\\alpha$} with $\\alpha$ fixed at 2.5 and with EBL attenuation approximated by N_obs = $e^{{-z/0.108}}$ N_intr; the deepest limits reach N(>0.5 TeV) ~ $10^{-12}$ $cm^{-2}$ $s^{-1}$. The paper's sharpest result concerns the radio galaxies M87 and IC 310: the HAWC limit on M87 is more than a factor of two below the 3FHL spectral extrapolation in the 0.5-2 TeV overlap and an order of magnitude below the 3FHL upper limit, and it is also below the 2005-2010 pointed-telescope measurements of M87 in high and intermediate states. For IC 310, the HAWC upper limit is about a factor of five below the hardest 3FHL point and below the reported flaring-state measurements. The conclusion drawn is that the persistent, time-averaged TeV output of these radio galaxies lies below the levels seen in activity-biased pointed observations.","pith_inferences":["A natural extension is to re-fit all 136 non-detections with a free spectral index and the full EBL optical depth rather than the fixed alpha = 2.5 exponential; the paper's own numbers show the redshift scale shifts from zh = 0.095 at alpha = 2.0 to zh = 0.118 at alpha = 3.0, so the tightest limits could move by tens of percent.","If the M87 baseline is genuinely this low, then monitoring campaigns that combine wide-field surveys with fast follow-up could measure the duty cycle of radio-galaxy TeV flares, i.e., the fraction of time these sources spend in high states, rather than just their peak outputs.","Stacking the non-detected AGN by redshift could turn this survey into a population-level probe of the EBL or of the average TeV luminosity function of nearby AGN, independently of any individual flaring episode."],"forward_implications":["If the HAWC upper limits are correct, then the vast majority of z <= 0.3 AGN have persistent N(>0.5 TeV) at or below the 10^-12 to 10^-11 cm^-2 s^-1 range, so any future TeV survey that wants to see quiescent AGN must reach below this flux.","For M87, the result implies its 0.5-2 TeV baseline is below the 3FHL extrapolated spectrum, meaning the catalog average is likely dominated by activity states rather than steady emission.","For IC 310, the hard 3FHL spectrum must soften or switch off in the TeV band during quiescence, since HAWC sees less flux than the hardest 3FHL point.","Continued HAWC exposure should either convert the best upper limits (M87, IC 310, 1ES 2344+514) into detections or push them deeper, distinguishing steady emitters from one-sided flaring sources."],"supporting_citations":[{"why":"Supplies the 138-object AGN sample and the 3FHL spectra to which HAWC limits are compared.","marker":"[16]"},{"why":"Provides the EBL attenuation model used to compute all flux limits and the exponential redshift approximation.","marker":"[15]"},{"why":"Validates HAWC's sensitivity and defines the fhit energy-binning from a 5-sigma-per-transit Crab detection.","marker":"[13]"},{"why":"Defines the standard HAWC analysis bins and flags bright nearby sources that could contaminate the survey.","marker":"[14]"},{"why":"Provides the confidence-interval prescription used to set 95% upper limits on non-detected sources.","marker":"[20]"},{"why":"Supplies the pointed-telescope high and intermediate state measurements of M87 that the HAWC limit is compared with.","marker":"[22]"},{"why":"Supplies the flaring-state measurements of IC 310 that the HAWC limit falls below.","marker":"[23]"},{"why":"Lists VHE AGN detections from pointed telescopes, used to identify additional TeV candidate sources in the sample.","marker":"[17]"}],"fun_headline_variants":["HAWC pins M87's TeV glow below pointed-telescope flares","TeV survey: only two blazars shine, M87 and IC 310 stay dark","HAWC's deep limits erase M87's steady TeV emission","After 1017 days, HAWC sees just two AGN in TeV light","M87 and IC 310: HAWC finds no persistent TeV emission"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Every non-detection upper limit assumes the source shines with a simple power-law spectrum of index 2.5 and that EBL absorption follows one exponential scale zh = 0.108; if any of these sources has a curved, harder, or softer spectrum, or if the EBL model is wrong, the limits and the M87 comparison shift.","fun_headline_variants_meta":{"raw":{"variants":["HAWC pins M87's TeV glow below pointed-telescope flares","TeV survey: only two blazars shine, M87 and IC 310 stay dark","HAWC's deep limits erase M87's steady TeV emission","After 1017 days, HAWC sees just two AGN in TeV light","M87 and IC 310: HAWC finds no persistent TeV emission"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000859,"raw_usage":{"total_tokens":3794,"prompt_tokens":1076,"completion_tokens":2718,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":692,"completion_tokens_details":{"reasoning_tokens":2613}},"tokens_in":692,"tokens_out":2718,"duration_ms":17102,"temperature":1.0,"reasoning_tokens":2613,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:33:41.875740+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Reanalyze the same 1017 days of HAWC data at the position of M87 with simulated source injections at the 3FHL-extrapolated flux: if the likelihood analysis cannot recover such a source at the claimed sensitivity, the upper limit is not a valid test of that extrapolation. Alternatively, a future wide-field TeV exposure with longer integration that measures M87's time-averaged N(>0.5 TeV) above 0.95 x $10^{-12}$ $cm^{-2}$ $s^{-1}$ would directly contradict the reported limit, assuming the source's baseline state is unchanged.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the 138-object AGN sample and the 3FHL spectra to which HAWC limits are compared."},{"cited_title":"2011, MNRAS 410, 2556","cited_arxiv_id":null,"evidence_quote":"Provides the EBL attenuation model used to compute all flux limits and the exponential redshift approximation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Validates HAWC's sensitivity and defines the fhit energy-binning from a 5-sigma-per-transit Crab detection."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the standard HAWC analysis bins and flags bright nearby sources that could contaminate the survey."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the confidence-interval prescription used to set 95% upper limits on non-detected sources."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the pointed-telescope high and intermediate state measurements of M87 that the HAWC limit is compared with."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the flaring-state measurements of IC 310 that the HAWC limit falls below."},{"cited_title":"of 30th ICRC 3, 1341","cited_arxiv_id":null,"evidence_quote":"Lists VHE AGN detections from pointed telescopes, used to identify additional TeV candidate sources in the sample."}],"review_version":1}