{"id":"d76e54c1-6fc2-4abe-aa47-36ac3e6e579a","arxiv_id":"1908.03112","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The hot coronae of NGC 4388 and NGC 2110 have temperatures near 75 to 80 keV and optical depth near 2, consistent with pair-production thermostats.","lead":"Astronomers combined X-ray data from three space telescopes to measure the temperature of the hot corona around two supermassive black holes. They found both coronae sit near the temperature where particle-antiparticle pairs form, supporting a thermostat model.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The joint fits tie the high-energy cutoff across a 13-year INTEGRAL average and several short XMM/NuSTAR snapshots; if the cutoff varies with flux or time, the reported kTe is a flux-weighted average rather than a single physical coronal temperature.","rationale":"The reader's verdict already identifies this as the weakest assumption, and my reading agrees. The paper's statistical fits are good, and the high-energy rollover is directly detected in IBIS (χ2/dof = 8/9 for cutoffpl versus 17/10 for a simple power law), so the measurement is not empty. However, the inference from a rollover in a 13-year average to a single physical kTe requires that the rollover energy be stable over the averaging window. The paper's own Appendix A shows a different spectral state in 2011, and Fedorova et al. (2011) reported cutoff variability, so this is not a hypothetical worry. The fix is straightforward: untie Ec among epochs and see whether the data demand variability. I would keep the CONDITIONAL verdict: the measurement is plausible and well-presented, but the central number should not be accepted as a single physical temperature until the epoch-resolved test is done. I do not see a more serious problem: the compps geometry and seed-temperature assumptions are explicit and at least partially tested, and the no-reflection result is consistent with previous work.","tokens_in":17760,"tokens_out":4807,"duration_ms":47862,"concrete_test":"Re-fit the broad-band spectra with Ec (and kTe/y in compps) untied among IBIS and each XMM/NuSTAR epoch, keeping all other model components as in Tables 4 and 5. If the epoch-by-epoch Ec values are mutually consistent at 90% confidence and the fit improvement is not significant (e.g. Δχ2 < 6 for three additional degrees of freedom for NGC 4388), the constant-cutoff assumption is supported and the concern is resolved. If the values disagree, the reported kTe is an average over spectral states and the paper's central claim should be downgraded or rephrased.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is that a single cutoff energy (and a single kTe/y in compps) describes both the 2003-2015 INTEGRAL/IBIS time-averaged spectrum and the individual XMM/NuSTAR snapshots. In Sects. 3.1.2 and 3.2.2 the model is tied across all datasets except for normalizations and, for NGC 2110, NH; the paper explicitly states that leaving NH, CF, or photon index free gives no significant improvement, but it does not report testing whether Ec itself varies between epochs. This matters because the central claim is a physical coronal temperature, not just a statistical description of the average spectrum. If the corona's temperature varies with flux or time, the IBIS average is a superposition of spectra with different rollovers, and a single-component fit returns a biased effective cutoff that may not correspond to any real kTe. The paper itself contains direct evidence that the assumption is questionable: Fedorova et al. (2011) reported cutoff variations between 80-100 keV and >320 keV for NGC 4388; the 2011 XMM3 observation is excluded for being a different spectral state (Appendix A); and the cross-normalization constants imply flux swings of factors ~3 (NGC 4388) and ~8 (NGC 2110). Since cutoff constancy is never tested, the reported kTe and the pair-thermostat comparison rest on an untested assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a spectral analysis of the hard X-ray (20-300 keV) INTEGRAL/IBIS time-averaged spectra of the Seyfert galaxies NGC 4388 and NGC 2110, combined with archival XMM-Newton and NuSTAR snapshots. The authors fit the broad-band spectra with xspec using an absorbed cutoff power law plus a Fe Kα line, obtaining well-constrained high-energy cutoffs Ec = 200+75-40 keV for NGC 4388 and Ec = 320+100-60 keV for NGC 2110, with no significant Compton reflection component (R < 0.12 and R < 0.02). Replacing the power law with the thermal Comptonisation model compps in spherical geometry yields coronal temperatures kTe = 80+40-20 keV (NGC 4388) and kTe = 75+20-15 keV (NGC 2110), with Compton parameters y ~ 1.1-1.2 corresponding to optical depths of roughly 1.7 and 2.1. The authors then estimate bolometric luminosities, compute the compactness parameter l = Lσ_T/(R m_e c^3) assuming R = 10 R_g, and argue that both objects fall below the pair runaway line in the l - Θ_e plane, consistent with the pair thermostat scenario.","tokens_in":17939,"tokens_out":10357,"duration_ms":97170,"significance":"If the measurements are correct, this work adds two well-constrained coronal temperature and optical depth measurements to the AGN sample, demonstrating the importance of high signal-to-noise data above 100 keV. The spectral analysis is statistically careful, the reduced spectra are made available, and the paper explicitly compares with and reconciles previous INTEGRAL, NuSTAR, Suzaku, and BeppoSAX results. The main quantitative conclusion (kTe ~ 75-80 keV) is, however, dependent on the assumption that a single cutoff energy describes both the 13-year INTEGRAL average and the short XMM/NuSTAR snapshots, and the pair-thermostat interpretation depends on an assumed coronal radius and bolometric correction. These dependencies are acknowledged in part but deserve further testing.","major_comments":[{"comment":"The high-energy cutoff Ec (and the equivalent kTe in the compps model) is tied across the time-averaged IBIS spectrum and the individual XMM-Newton and NuSTAR snapshots, and the paper does not test whether Ec varies between epochs. The cross-normalization constants imply flux variations of factors ~3 for NGC 4388 and ~8 for NGC 2110, and Fedorova et al. (2011) reported possible cutoff variations in NGC 4388 between 80-100 keV and >320 keV. If the cutoff varies with flux or time, the reported Ec and kTe are flux-weighted effective values rather than a single physical coronal temperature. I recommend allowing Ec (or kTe) to vary among the spectra in the joint fit, or splitting the IBIS data into temporal subsets, to demonstrate that the cutoff is stable; otherwise the central claim of a measured coronal temperature is not fully supported.","section":"Secs. 3.1.2 and 3.2.2"},{"comment":"The pair-thermostat consistency check depends on the assumed corona radius through l = Lσ_T/(R m_e c^3), and the paper assumes R = 10 R_g with a bolometric correction from Marconi et al. (2004). The resulting compactness differs by a factor of about 26 between the two sources, and for NGC 2110 l ~ 0.5 at R = 10 R_g, which is far below the pair runaway line. This makes the statement of 'consistency with the pair thermostat' relatively weak; the authors should explicitly discuss the sensitivity of the compactness and of the conclusion to the assumed radius and luminosity, and present the result as an illustrative consistency check rather than a quantitative confirmation.","section":"Sec. 4"},{"comment":"The 2011 XMM-Newton observation (XMM3) of NGC 4388 is excluded from the joint fit because it likely represents a different spectral state (flatter photon index and lower column density), yet this epoch is included in the IBIS time average used in the same joint fit. The authors should clarify whether this exclusion is consistent with their assumption of a single cutoff across the full 2003-2015 IBIS average, or whether the inclusion of the 2011 state could bias the time-averaged cutoff and hence the derived kTe.","section":"Sec. 3.1.2 and Appendix A"}],"minor_comments":[{"comment":"The cross-normalization constants in Tables 4 and 5 (e.g., K_IBIS-pn, K_IBIS-NusA) are not fully defined in the text; a sentence explaining that they are multiplicative factors between the IBIS and XMM/NuSTAR normalizations would improve readability.","section":"Sec. 3.1.2"},{"comment":"The unabsorbed 0.1-200 keV luminosity used in the compactness calculation is not reported; giving this value (with its uncertainty) would make the l - Θ_e placement reproducible.","section":"Sec. 4"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational contribution and the data release is appreciated. The main risk is the untested constancy of the high-energy cutoff across epochs; if the authors show that allowing Ec (or kTe) to vary does not change their conclusions, the paper would be suitable for publication. The pair-thermostat discussion should be reframed as an illustrative consistency check given the radius assumption."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know before you spend time here. First, this is a clean measurement paper, not a headline-grabber: two bright Seyferts get well-constrained high-energy cutoffs from 13 years of INTEGRAL/IBIS, with XMM-Newton and NuSTAR anchoring the soft band. Second, the genuinely new result is the NGC 2110 coronal temperature: kTe = 75(+20,-15) keV and a 320(+100,-60) keV cutoff, which revises the much higher value reported by Lubiński et al. (2016) and replaces earlier lower limits. That is worth having in the census. The analysis is solid. The cutoff power-law and compps fits are statistically good; the reflection component is constrained to be negligible; the reduced spectra appear to be available. The authors also directly test the source of the Lubiński discrepancy by forcing R=0.6, which drives kTe to about 340 keV with a worse fit. That is a concrete, falsifiable explanation. The exclusion of the 2011 XMM3 observation is documented in an appendix, not hidden. No circularity in the main argument: kTe and tau are fit outputs, and the pair-thermostat comparison is an external consistency check. The caveat that actually matters is the one the stress-test flags: the joint fit ties the cutoff and kTe across the 2003-2015 IBIS average and every snapshot, freeing only normalizations and NH. The authors say freeing NH, CF, or photon index makes no difference; they never test freeing the cutoff. If the rollover moves with flux or time, the IBIS average smears it into an effective cutoff that is not a single physical coronal temperature. The paper itself cites Fedorova et al. (2011) reporting cutoff variability in NGC 4388, and the XMM3 exclusion is evidence that spectral states do differ. I would not call this fatal. The low-energy snapshots are broadly consistent with the time-averaged shape, and the paper's own conclusions are phrased as consistency, not proof. But the kTe values carry an extra systematic uncertainty that is not in the error bars, and an epoch-resolved test would settle it. The compactness-temperature agreement also assumes R=10 gravitational radii plus a bolometric correction; standard, but illustrative rather than a test. Bottom line: observational AGN people will use these numbers, and the paper deserves peer review and likely publication. I would cite it for the NGC 2110 measurement. Bring it to a reading group if you want a clean example of how to resolve a literature discrepancy with better broad-band data.","headline":"Clean spectral measurement that gives NGC 2110 a credible coronal temperature and revises an inflated earlier value; the main caveat is an untested assumption that the high-energy cutoff is constant across 13 years of INTEGRAL and snapshot data.","tokens_in":18608,"tokens_out":5350,"would_cite":true,"duration_ms":52530,"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":"Two Seyfert coronae are measured at 75–80 keV, placing them in the pair-thermostat regime.","keywords":["active galactic nuclei","Seyfert galaxies","X-ray corona","coronal temperature","Comptonisation","high-energy cutoff","INTEGRAL","pair production"],"falsifier":"Split the INTEGRAL data into epochs matching the XMM–Newton and NuSTAR snapshots and fit the cutoff freely in each epoch; if the derived cutoffs move outside the reported 90% confidence ranges ($200^{+75}_{-40}$ keV for NGC 4388, $320^{+100}_{-60}$ keV for NGC 2110), the time-averaged temperature is an artefact of spectral variability and the pair-thermostat interpretation fails.","tokens_in":17454,"feed_emoji":"🌌","tokens_out":16470,"duration_ms":139227,"temperature":0.7,"pith_summary":"This paper uses 13 years of INTEGRAL hard X-ray data (20–300 keV) joined with archival XMM-Newton and NuSTAR spectra to pin down the high-energy cutoff in two bright Seyfert galaxies, NGC 4388 and NGC 2110. Interpreting the cutoff with a thermal Comptonisation model, it derives coronal temperatures of about 80 keV and 75 keV for the two sources, with optical depth near 2. These values place both sources in the regime where electron–positron pair production is expected to act as a thermostat, capping the coronal temperature. The result matters because direct measurements of coronal temperatures above 100 keV are rare, and the pair-thermostat scenario is a key prediction of how AGN coronae regulate themselves.","feed_headline":"Two Seyfert coronae measured at 75–80 keV: pairs act as thermostat","feed_subtitle":"A 13-year INTEGRAL baseline reveals the high-energy cutoff that sets the coronal temperature.","key_machinery":"The analysis rests on the thermal Comptonisation model compps (Poutanen & Svensson 1996), which computes the spectrum produced when soft disc photons are upscattered by a hot electron plasma; fitting the electron temperature $k T_{\\rm e}$ and the Compton parameter $y=4\\tau\\,k T_{\\rm e}/m_{\\rm e}c^{2}$ (used instead of $\\tau$ to reduce the known temperature–optical-depth degeneracy) yields the coronal parameters under an assumed spherical geometry. The load-bearing data are the INTEGRAL/IBIS spectra accumulated from 2003 to 2015, whose coverage up to 300 keV makes the high-energy turnover visible; the turnover is modelled both as an exponential cutoff and as the intrinsically sharper Comptonisation rollover. The compactness–temperature diagram ($\\ell$ vs $\\Theta_{\\rm e}=k T_{\\rm e}/m_{\\rm e}c^{2}$), with the pair runaway line of Fabian et al. (2015), is used to interpret the measured temperatures.","core_discovery":"Joint fitting of the time-averaged INTEGRAL/IBIS spectrum (20–300 keV) with archival XMM–Newton and NuSTAR spectra shows that both sources are well described by an absorbed cutoff power law with no Compton reflection component. The high-energy cutoff is measured at $E_{\\rm c}=200^{+75}_{-40}$ keV for NGC 4388 and $E_{\\rm c}=320^{+100}_{-60}$ keV for NGC 2110. Replacing the phenomenological cutoff with the thermal Comptonisation model compps in spherical geometry gives coronal temperatures of $k T_{\\rm e}=80^{+40}_{-20}$ keV (NGC 4388) and $k T_{\\rm e}=75^{+20}_{-15}$ keV (NGC 2110), with optical depths $\\tau \\simeq 1.7$ and $\\tau \\simeq 2.1$. The derived temperatures and luminosities place both objects below the pair runaway line in the compactness–temperature plane, supporting the pair-thermostat scenario in which electron–positron pair production regulates the coronal temperature.","pith_inferences":["If the pair-thermostat picture is right, the coronal temperature of these two sources should remain near 75–80 keV while their luminosity varies by a factor of several; a direct test would be to split the INTEGRAL data into high- and low-flux epochs and measure $k T_{\\rm e}$ in each.","The two galaxies have very different Eddington ratios (~0.23 and ~0.01) yet nearly identical coronal temperatures; this hints that $k T_{\\rm e}$ is not driven by accretion rate in this regime, a correlation that a larger sample of sources with measured black hole masses could test.","Because Fedorova et al. (2011) suggested the cutoff in NGC 4388 varies between about 80–100 keV and above 320 keV, the time-averaged 200 keV cutoff may be a blend of states; if so, the single-temperature corona would need to be replaced by a distribution of temperatures, and the pair-thermostat interpretation would need to be revisited."],"forward_implications":["Assuming a coronal radius of about 10 gravitational radii, both sources lie below the pair runaway line in the $\\ell$–$\\Theta_{\\rm e}$ plane, so pair production can cap the coronal temperature as the pair-thermostat model predicts.","The lack of a Compton reflection hump alongside strong Fe K$\\alpha$ lines indicates the line arises in Compton-thin material rather than from a disc reflection continuum.","The measured cutoffs of 200 and 320 keV are high relative to the typical INTEGRAL cutoff (mean ~128 keV), and they place both sources in the small group of AGNs with coronal temperatures above 70 keV that are measured to better than about 50 per cent uncertainty.","Within the epochs used, the primary continuum slope is consistent with being constant across the 2003–2015 INTEGRAL average and the XMM–Newton/NuSTAR snapshots, with the 2011 XMM observation of NGC 4388 as a distinct flatter state."],"supporting_citations":[{"why":"Supplies the compps thermal Comptonisation model used to convert the observed spectrum into coronal temperature and optical depth.","marker":"Poutanen & Svensson 1996"},{"why":"Defines the compactness–temperature diagram and the pair runaway line used to place both sources in the pair-thermostat regime.","marker":"Fabian et al. 2015"},{"why":"Provides the NuSTAR spectrum of NGC 4388 and the partial-covering absorption model, and reported no Compton reflection hump.","marker":"Kamraj et al. 2017"},{"why":"Provides the NuSTAR spectrum of NGC 2110, the lower limit on its cutoff, and the variable Fe Kα line decomposition.","marker":"Marinucci et al. 2015"},{"why":"Reports the previous INTEGRAL coronal temperatures (53 and 230 keV) that this paper refines and, for NGC 2110, revises downward.","marker":"Lubiński et al. 2016"},{"why":"Reported variability of the NGC 4388 high-energy cutoff between about 80–100 keV and above 320 keV, the main challenge to the constant-cutoff assumption.","marker":"Fedorova et al. 2011"},{"why":"Earlier INTEGRAL measurement of the NGC 4388 cutoff (about 202 keV) that is confirmed here.","marker":"Molina et al. 2013"},{"why":"Early INTEGRAL+XMM analysis of NGC 4388 giving a cutoff lower limit and the absorbing column density used in the fit.","marker":"Beckmann et al. 2004"}],"fun_headline_variants":["INTEGRAL pins two Seyfert coronae at ~75–80 keV","AGN coronae at 75–80 keV: pair production as thermostat","Coronal temperature measured: 75–80 keV in two AGNs","Pair thermostat confirmed: coronae of NGC 4388 & 2110 at ~80 keV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes the intrinsic photon index and high-energy cutoff stayed constant across the 2003–2015 INTEGRAL average and the XMM–Newton and NuSTAR snapshots, so the measured cutoff represents a single physical corona.","fun_headline_variants_meta":{"raw":{"variants":["INTEGRAL pins two Seyfert coronae at ~75–80 keV","AGN coronae at 75–80 keV: pair production as thermostat","Coronal temperature measured: 75–80 keV in two AGNs","Pair thermostat confirmed: coronae of NGC 4388 & 2110 at ~80 keV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000838,"raw_usage":{"total_tokens":3673,"prompt_tokens":985,"completion_tokens":2688,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":601,"completion_tokens_details":{"reasoning_tokens":2602}},"tokens_in":601,"tokens_out":2688,"duration_ms":22848,"temperature":1.0,"reasoning_tokens":2602,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:23:46.290657+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Split the INTEGRAL data into epochs matching the XMM–Newton and NuSTAR snapshots and fit the cutoff freely in each epoch; if the derived cutoffs move outside the reported 90% confidence ranges ($200^{+75}_{-40}$ keV for NGC 4388, $320^{+100}_{-60}$ keV for NGC 2110), the time-averaged temperature is an artefact of spectral variability and the pair-thermostat interpretation fails.","supporting_citations":[{"cited_title":"A., Brightman , M., & Balokovi \\'c , M","cited_arxiv_id":null,"evidence_quote":"Provides the NuSTAR spectrum of NGC 4388 and the partial-covering absorption model, and reported no Compton reflection hump."},{"cited_title":"V., Beckmann , V., Neronov , A., & Soldi , S","cited_arxiv_id":null,"evidence_quote":"Reported variability of the NGC 4388 high-energy cutoff between about 80–100 keV and above 320 keV, the main challenge to the constant-cutoff assumption."}],"review_version":1}