{"id":"02fef49f-c5f3-420d-894d-dcc46eec14a0","arxiv_id":"1909.01179","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"HAWC observations show the intrinsic spectrum of Mrk 421 requires an exponential cutoff near 5 TeV, while Mrk 501 is consistent with a power law up to 100 TeV.","lead":"This paper reports time-averaged very-high-energy spectra of two bright blazars, Mrk 421 and Mrk 501, measured with HAWC over 837 days. The authors find that Mrk 421 shows a spectral cutoff near 5 TeV after correcting for background light, while Mrk 501 is described by a power law.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Mrk 421 cutoff claim lacks reported model-comparison significance; without a ΔTS between the PL and PL+CO fits, the central spectral difference is not established.","rationale":"The reader's weakest_assumption centered on EBL model dependence. That is a legitimate uncertainty, but it is not the most load-bearing condition for the central claim. The central claim is a model-selection statement: Mrk 421 requires a cutoff, Mrk 501 does not. The paper never reports the likelihood difference between the two spectral models for either source, so the reader cannot verify that the cutoff is preferred at any significance. This is a direct gap in the evidence for the headline result, not merely a matter of systematic choices. It is also more specific than the EBL concern: if ΔTS is small, the cutoff is unsupported regardless of which EBL model is used. For Mrk 501, the claim 'without requiring an energy cut-off' is an overstatement when the fit cannot constrain Ec; the paper itself notes the PL and PL+CO fits have very similar TS values and Ec > 700 TeV, which means the cutoff parameter is effectively unbounded, not measured to be absent. The conditional verdict remains appropriate: the paper is a clear preliminary analysis, but the central spectral comparison needs an explicit likelihood-ratio statistic and a treatment of energy-scale systematics before the cutoff claim can be accepted. My proposed check targets the missing statistic directly and also tests EBL robustness, addressing both the primary gap and the reader's concern.","tokens_in":4170,"tokens_out":4128,"duration_ms":43872,"concrete_test":"Re-run the same LiFF global fits for Mrk 421 using equations 3.2 and 3.3 with identical energy bins, and report ΔTS = TS(PL+CO) − TS(PL) along with the profile-likelihood 68% and 95% confidence intervals for Ec. If ΔTS is below about 9 (roughly 3σ for one additional degree of freedom) or the 95% interval for Ec extends above about 10 TeV, the cutoff claim in Table 1 is not statistically supported. As a robustness check, repeat the comparison using the Domínguez et al. (2011) EBL model instead of Gilmore et al. (2012) to see whether Ec shifts by more than the quoted 1.02 TeV statistical uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the intrinsic spectrum of Mrk 421 is better described by a power law with an exponential cutoff (Ec = 5.24 ± 1.02 TeV) while Mrk 501 is consistent with a pure power law. The only quantitative support is the statement that 'the model that best describes the intrinsic spectrum' is PL+CO for Mrk 421 and that for Mrk 501 the PL and PL+CO fits have 'very similar' TS values with a fitted cutoff above 700 TeV. No ΔTS between the two models is reported. Because PL is nested in PL+CO (the cutoff model reduces to a power law as Ec → ∞), the natural test is a likelihood-ratio comparison: ΔTS = TS(PL+CO) − TS(PL). Without this number, a genuine cutoff cannot be distinguished from a mild preference or an unconstrained high-energy tail. The quoted statistical uncertainties on Ec also ignore HAWC's energy-scale systematics in exactly the 1–10 TeV region where the cutoff is claimed. The EBL model (Gilmore et al. 2012) is a secondary concern, but since Mrk 421 and Mrk 501 are at nearly equal redshift, an EBL error would affect both spectra in a similar way and is less likely to invert the relative conclusion; the missing model-comparison statistic can directly invalidate the cutoff claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a time-averaged spectral analysis of the blazars Mrk 421 and Mrk 501 using 837 days of HAWC data (June 2015 to December 2017). The authors fit power-law (PL) and power-law-with-exponential-cutoff (PL+CO) models to counts in energy bins above 1 TeV, applying a forward-folding maximum-likelihood procedure and correcting the intrinsic spectrum for extragalactic background light (EBL) absorption using the Gilmore et al. (2012) model. They conclude that the intrinsic spectrum of Mrk 421 is better described by a PL+CO with photon index 2.22 ± 0.10 and cutoff energy 5.24 ± 1.02 TeV, while Mrk 501 is consistent with a single power law of index 2.40 ± 0.06 without a cutoff.","tokens_in":4476,"tokens_out":5061,"duration_ms":50492,"significance":"If the results hold, this would be one of the first HAWC measurements of the intrinsic VHE spectra of these two archetypal high-synchrotron-peaked BL Lacs above 1 TeV, and it would suggest a spectral difference between them that could bear on blazar emission models. The analysis uses standard procedures: the HAWC maximum-likelihood framework, a published energy estimator, and a widely used EBL model. The paper is transparent about using statistical uncertainties only and about its preliminary character. However, the central claim of curvature in Mrk 421 and its absence in Mrk 501 is not statistically quantified, and systematic uncertainties—especially energy-scale and EBL-model dependence—are not addressed. Because the model comparison statistic is missing, the significance of the paper's main conclusion cannot currently be assessed.","major_comments":[{"comment":"The central claim that Mrk 421 is better described by PL+CO rests on a model comparison that is never reported. Since the power law is nested in PL+CO (Ec → ∞), the likelihood-ratio test statistic ΔTS = TS(PL+CO) − TS(PL) is the appropriate measure, and the paper should quote it for both sources. Without ΔTS, the reported Ec = 5.24 ± 1.02 TeV for Mrk 421 cannot be distinguished from a marginal preference for an extra parameter, and the statement that Mrk 501's fits have 'very similar' TS values is unverifiable. I request the ΔTS values, and ideally a profile likelihood in Ec, for both sources.","section":"§4, Table 1"},{"comment":"The uncertainties quoted in Table 1 are statistical only, but the key parameter, Ec = 5.24 ± 1.02 TeV, lies in the energy range where HAWC's energy-scale systematics are most relevant, given the energy estimator described in Section 2 and reference [10]. The authors should either estimate and propagate systematics from the detector response and energy scale or at least bound their effect on Ec. In addition, the EBL model of Gilmore et al. (2012) is used without testing alternatives; because the intrinsic cutoff is obtained by dividing the observed spectrum by exp(−τ), repeating the fits with another EBL model (e.g., Domínguez et al. 2011 or Franceschini et al. 2008) would demonstrate whether the Mrk 421 cutoff is robust or an artifact of the chosen τ(E).","section":"§4, Table 1"},{"comment":"The treatment of Mrk 501's cutoff is not quantitatively supported. The text states that the PL and PL+CO fits have 'very similar' TS values and that the fitted cutoff is larger than 700 TeV, but no numbers are reported, and the value '∞' in Table 1 is not a fit result. The authors should give the fitted Ec for Mrk 501 with its uncertainty, or a lower limit, and the ΔTS between PL and PL+CO; this is necessary to justify the claim that Mrk 501 is 'well described' by a power law without requiring a cutoff. As written, the comparison between the two sources is not established.","section":"§4, Table 1"}],"minor_comments":[{"comment":"Equations (3.2) and (3.3) are labeled 'PL' and 'PL+CO', but they include the EBL attenuation factor exp(−τ). Since the text says the input model is the intrinsic one and is then attenuated, these equations should be clarified as the observed (attenuated) models actually compared with the data, with the intrinsic models written without exp(−τ).","section":"§3, Eqs. (3.2) and (3.3)"},{"comment":"The column header '√TS' is ambiguous; the table should state that this is the detection significance in units of σ, not a spectral-fit test statistic.","section":"Table 1"},{"comment":"The figure caption should explain how upper limits are plotted, since Section 3 states that bins with TS < 25 are replaced by 95% confidence upper limits; the caption currently does not mention this.","section":"Figure 1"},{"comment":"The results are repeatedly called 'preliminary' (Abstract, Section 4, Section 5). If this is intended as a journal publication, the authors should either present the analysis as final or explicitly state which aspects remain preliminary (e.g., systematic uncertainties).","section":"Abstract, §4, §5"},{"comment":"The phrase 'photon-photon attenuation inside the source' is speculative; since EBL attenuation has already been removed, the residual curvature could also be described as intrinsic spectral curvature. Suggest rephrasing to avoid confusion.","section":"§5"}],"recommendation":"major_revision","confidential_remarks":"This is an ICRC proceedings paper whose central spectral-difference claim is not yet statistically supported. The main missing element is the ΔTS between the PL and PL+CO fits for both sources, which is straightforward to compute with the existing LiFF framework. I recommend that the editors require this addition, along with at least a discussion of systematic uncertainties, before accepting the paper for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a short HAWC proceedings paper with a solid measurement and a headline claim that outruns the statistics. The genuinely new thing is the time-averaged HAWC spectra of Mrk 421 and Mrk 501 above 1 TeV from 837 days of data, produced with the collaboration's standard forward-folding likelihood (LiFF) and corrected for EBL with Gilmore et al. The fitted parameters are plausible, and the paper is honest that these are preliminary results.\n\nWhat it does well: the measurement itself is reproducible in principle, the method is standard, and the flux points are useful as a cross-check to IACT results. The authors correctly separate observed and intrinsic spectra and flag that the EBL model enters the derivation.\n\nThe soft spot is the central spectral comparison. The abstract and summary claim Mrk 421 is \"better described\" by a power law with an exponential cutoff and Mrk 501 by a plain power law. But no ΔTS between the two nested models is ever reported. PL is nested in PL+CO (the latter reduces to PL as Ec→∞), so the natural test is a likelihood-ratio comparison. Without that number, the cutoff claim is not established: it could be a mild preference or an unconstrained high-energy tail. For Mrk 501 the fitted cutoff is above 700 TeV, which is fine, but again no uncertainty or ΔTS is given. Also, the quoted errors are statistical only; HAWC's energy-scale systematics in the 1–10 TeV region matter for a cutoff at about 5 TeV, and they are not discussed. The EBL model is a secondary concern; since both sources are at nearly the same redshift, an EBL error shifts both spectra in a similar way, so the relative conclusion is less vulnerable.\n\nThis is not a fatal flaw for a conference proceedings, but it is a load-bearing omission. If the paper is offered to a journal, it deserves a serious referee, who should ask for the model-comparison statistic, systematic uncertainties, and wording that matches the evidence (\"consistent with\" rather than \"better described\"). For the ICRC audience it is acceptable as a preliminary report, but I would not cite it as evidence for an intrinsic cutoff in Mrk 421.","headline":"Solid HAWC spectra of two blazars, but the claimed Mrk 421 cutoff is not statistically established because no ΔTS between nested models is reported.","tokens_in":5000,"tokens_out":2808,"would_cite":false,"duration_ms":26659,"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 intrinsic TeV spectra of Mrk 421 and Mrk 501 differ: Mrk 421 cuts off near 5.2 TeV, while Mrk 501 does not.","keywords":["Markarian 421","Markarian 501","blazar spectra","BL Lac objects","very high energy gamma rays","extragalactic background light","HAWC observatory","spectral cutoff"],"falsifier":"Re-fit the same HAWC data with an alternative EBL model whose optical depth at 5 TeV differs substantially from the adopted one: if Mrk 421's fitted cutoff moves outside roughly 3-8 TeV or disappears, the claim of an intrinsic ~5.2 TeV cutoff is weakened. A longer exposure that reveals a break in Mrk 501's spectrum above 10 TeV would falsify the pure-power-law description.","tokens_in":4012,"feed_emoji":"🔭","tokens_out":15698,"duration_ms":136745,"temperature":0.7,"pith_summary":"Using 837 days of data from the HAWC gamma-ray observatory, this paper tries to establish that the two nearest bright blazars have genuinely different intrinsic TeV spectra once absorption by the extragalactic background light is removed: Markarian 421 rolls over exponentially near 5.2 TeV, while Markarian 501 is consistent with a pure power law across the measured range. If the difference is real, it means that similar-looking objects can accelerate or radiate particles differently at the highest energies, giving a direct observational anchor for models of blazar jets. Because Mrk 421 and Mrk 501 are the reference blazars of the TeV sky, fixing their intrinsic shapes also sharpens multi-wavelength and emission-model comparisons built around them.","feed_headline":"Mrk 421 cuts off at 5 TeV; Mrk 501 does not","feed_subtitle":"HAWC's 837-day spectra show the two nearest bright blazars differ in intrinsic TeV shape.","key_machinery":"The machinery that carries the argument is forward-folding spectral fitting: a candidate intrinsic model is multiplied by the EBL attenuation $\\exp(-\\tau(E,z))$, convolved with the HAWC detector response, and compared with binned photon counts through a Poisson maximum-likelihood statistic $TS = 2\\ln(L_1/L_0)$. Event energies come from a ground-parameter estimator that uses the charge measured 40 m from the shower axis and divides the data into quarter-decade energy bins from 0.316 to 100 TeV; fits are restricted to above 1 TeV because the energy resolution is poor below that. The two candidate shapes are a single power law and a power law with an exponential cutoff, and the preferred shape is selected by comparing the resulting global-fit test statistics.","core_discovery":"After forward-folding candidate models through the detector response, the paper's central finding is that the intrinsic spectrum of Mrk 421 is best described by $dN/dE = N_0 (E/E_0)^{-2.22 \\pm 0.10} \\exp(-E/(5.24 \\pm 1.02~\\mathrm{TeV}))$, while Mrk 501 is well described by a single power law with index $2.40 \\pm 0.06$ and no cutoff. The EBL attenuation factor $\\exp(-\\tau)$ from the adopted model is applied before fitting, so the quoted parameters and the difference between the sources are presented as intrinsic. The detection significance for each source is expressed through $TS = 2\\ln(L_1/L_0)$, with $\\sqrt{TS}=48$ for Mrk 421 and $\\sqrt{TS}=22$ for Mrk 501. The paper reports these as preliminary results.","pith_inferences":["The paper leaves open whether the Mrk 421 cutoff is set inside the jet or by photon-photon absorption near the source; one testable extension is to watch whether the cutoff energy shifts during flaring activity, which would track the compactness of the emission region.","Applying the same EBL-corrected analysis to other high-synchrotron-peaked blazars could reveal whether a cutoff near a few TeV is a common feature; if it is, it would point to a shared limit on particle acceleration in these jets.","Repeating the fits with alternative EBL models is a direct robustness check: the same data could either confirm the 5.2 TeV cutoff as intrinsic or expose it as partly an artifact of the adopted absorption."],"forward_implications":["Mrk 421's emission model must reproduce an exponential cutoff at a few TeV, not just a continuation of the power law.","Mrk 501's intrinsic spectrum can be treated as a pure power law across the HAWC energy range, so the two blazars cannot share a single simple spectral template.","Since both sources have nearly equal redshifts, the reported difference is not mainly a distance or EBL effect under the adopted EBL model.","The fitted index and cutoff values provide concrete quantitative anchors for leptonic and hadronic jet models and for comparisons with the synchrotron component."],"supporting_citations":[{"why":"Supplies the EBL opacity tau(E,z) that multiplies each candidate intrinsic model by exp(-tau); all reported intrinsic parameters and the Mrk 421 cutoff depend on it.","marker":"[13]"},{"why":"Provides the ground-parameter energy estimator and the quarter-decade energy binning used to assign gamma-ray energies and to compute flux points.","marker":"[10]"},{"why":"Supplies the binned Poisson maximum-likelihood forward-folding fitter that compares candidate spectra with the observed HAWC counts.","marker":"[11]"},{"why":"Justifies the conversion of the likelihood-ratio test statistic into detection significance as sqrt(TS).","marker":"[12]"},{"why":"Provides the detector performance and the sigma > 5 detection criterion that define which sources are included.","marker":"[9]"}],"fun_headline_variants":["HAWC sees Mrk 421 break at 5 TeV, Mrk 501 smooth","TeV spectra split: Mrk 421 cuts, Mrk 501 doesn't","Blazar pair diverges: cutoff for 421, none for 501","Mrk 421 hits TeV wall; Mrk 501 sails through","HAWC pinpoints Mrk 421's 5 TeV cutoff, not 501"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the adopted extragalactic background light model gives the correct absorption: the observed spectrum is divided by that model's predicted attenuation before fitting, so a different EBL opacity would change the fitted intrinsic indices and especially the Mrk 421 cutoff.","fun_headline_variants_meta":{"raw":{"variants":["HAWC sees Mrk 421 break at 5 TeV, Mrk 501 smooth","TeV spectra split: Mrk 421 cuts, Mrk 501 doesn't","Blazar pair diverges: cutoff for 421, none for 501","Mrk 421 hits TeV wall; Mrk 501 sails through","HAWC pinpoints Mrk 421's 5 TeV cutoff, not 501"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000894,"raw_usage":{"total_tokens":3858,"prompt_tokens":951,"completion_tokens":2907,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":567,"completion_tokens_details":{"reasoning_tokens":2807}},"tokens_in":567,"tokens_out":2907,"duration_ms":18772,"temperature":1.0,"reasoning_tokens":2807,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:24:41.719636+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-fit the same HAWC data with an alternative EBL model whose optical depth at 5 TeV differs substantially from the adopted one: if Mrk 421's fitted cutoff moves outside roughly 3-8 TeV or disappears, the claim of an intrinsic ~5.2 TeV cutoff is weakened. A longer exposure that reveals a break in Mrk 501's spectrum above 10 TeV would falsify the pure-power-law description.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the binned Poisson maximum-likelihood forward-folding fitter that compares candidate spectra with the observed HAWC counts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Justifies the conversion of the likelihood-ratio test statistic into detection significance as sqrt(TS)."}],"review_version":1}