{"id":"15f9f209-6ea8-4d3b-82e0-1bbb7875e771","arxiv_id":"1908.02885","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A MaNGA-based sample of 152 AGN, combined with quasar data, yields an extended narrow-line region size-luminosity relation with slope 0.42 ± 0.02 over four orders of magnitude.","lead":"This paper uses 152 AGN from the MaNGA survey to measure how the size of the ionized narrow-line region scales with AGN luminosity. The authors find a slope of 0.42 ± 0.02 over four decades in luminosity, consistent with photoionization by the central AGN rather than jets or outflows.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The combined fit does not explicitly state whether the 10^-16 surface-brightness threshold is rest-frame; if applied in the observed frame, the Liu et al. quasar sizes are systematically too large and the slope 0.42±0.02 is biased.","rationale":"The reader's weakest_assumption correctly identifies the most load-bearing uncertainty: the equivalence of the 10^-16 surface-brightness threshold across the MaNGA and Liu et al. samples in the presence of cosmological dimming. I confirmed that the paper's Section 3.3 raises the (1+z)^4 issue but Section 4, when describing the quasar data and remeasurement of R16, does not explicitly say the threshold was converted to rest frame. This directly affects the combined fit in Eq. 2 and Figure 6, which is the central assertion. The concern is concrete and testable: one can inspect the Liu et al. methodology and re-run the fit under the rest-frame convention. Because the reader's report already conditionally accepts the paper and this concern is precisely the stated condition, no verdict change is needed. I do not see another more fundamental flaw; the paper's PSF treatment, sample selection, and fitting procedures are described in reasonable detail for an IFU study, and the slope from MaNGA alone (0.49 ± 0.04) is a useful internal check. The main numerical claim should therefore remain CONDITIONAL pending clarification of the dimming convention in the combined sample.","tokens_in":26450,"tokens_out":6349,"duration_ms":69164,"concrete_test":"Check Liu et al. (2013, 2014) to determine whether their published surface-brightness profiles and thresholds are observed-frame or rest-frame. Then, for each quasar, remeasure R16 at a rest-frame threshold of 10^-16 erg/s/cm^2/arcsec^2 (multiply observed SB by (1+z)^4 before interpolation) and refit Eq. 2 with the same Kelly (2007) Bayesian procedure used in the paper. If the slope changes by more than ~0.05 relative to the published 0.42 ± 0.02, the central claim is sensitive to the dimming convention and the paper must explicitly state and correct for it.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Eq. 2, Fig. 6) rests on merging MaNGA Seyferts with higher-redshift Liu et al. (2013, 2014) quasars using a common isophotal threshold, R16 at 10^-16 erg/s/cm^2/arcsec^2. Section 3.3 warns that cosmological dimming changes surface brightness by (1+z)^4, but Section 4 does not state that the Liu et al. profiles were converted to rest frame before R16 was measured. For z = 0.5 the dimming factor is about 5; for z = 1 it is about 16. If the threshold is applied in the observed frame for the quasars while the MaNGA threshold is effectively rest-frame (z ~ 0.03, correction ~1.1), quasar R16 values correspond to a much lower intrinsic surface brightness and are systematically overestimated. The fit then combines sizes defined at different intrinsic thresholds. The quoted slope is already flatter than the MaNGA-only value 0.49 ± 0.04, so the discrepancy could be tied to this convention. The direction and magnitude depend on the actual quasar redshifts and profile shapes, but the paper does not document which convention is used. This is a load-bearing, testable ambiguity, not a proven error: if Liu et al. already quote rest-frame thresholds, the concern disappears.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper measures the size-luminosity relation of extended narrow line regions (ENLRs) in a sample of 152 MaNGA AGN selected via spatially resolved BPT diagnostics. ENLR sizes are defined by a 10^-16 erg/s/cm^2/arcsec^2 [O III] surface brightness isophote (R16), measured from PSF-deconvolved Sersic fits to the [O III] surface brightness profiles. By combining the MaNGA Seyferts with IFU quasar data from Liu et al. (2013, 2014), the authors fit a log-linear relation, log(R16/pc) = (0.42 +/- 0.02) log(L[O III]/erg/s) - (13.97 +/- 0.95), over four orders of magnitude in luminosity (Eq. 2, Fig. 6). They interpret the slope with the Dempsey & Zakamska (2018) photoionized cloud model and argue that outflows and jets are not required to explain the ENLR extension in low-luminosity Seyferts.","tokens_in":26754,"tokens_out":8740,"duration_ms":98196,"significance":"If the central result holds, it provides a well-calibrated ENLR size-luminosity relation over a much wider luminosity range than previous studies, connecting local Seyferts to luminous quasars with a uniform IFU-based size definition. The analysis has notable strengths: Monte Carlo radius uncertainties, PSF-deconvolved Sersic profile fitting, dust correction via the Balmer decrement, and a Bayesian regression method (Kelly 2007) are used, and the code is publicly available. The main caveats concern the rest-frame versus observed-frame threshold convention and the treatment of unresolved galaxies/upper limits; these issues are testable and should be clarified before the quoted slope is taken at face value.","major_comments":[{"comment":"The paper never states whether the 10^-16 erg/s/cm^2/arcsec^2 surface brightness threshold is defined in the rest frame or the observed frame. Section 3.3 notes that cosmological dimming scales surface brightness by (1+z)^4 and says this is important for high-redshift comparisons, but Section 4 does not specify that the Liu et al. (2013, 2014) profiles were converted to rest frame before measuring R16. For z ~ 0.5-1 quasars the dimming factor is 5-16, so an observed-frame threshold would make the quasar R16 values larger than a rest-frame threshold would, which would flatten the combined slope relative to the MaNGA-only value of 0.49 +/- 0.04. Please state the convention explicitly, verify what Liu et al. actually quote, and re-fit with the threshold consistently defined (or demonstrate that the two conventions give the same result at the relevant redshifts).","section":"Section 3.3 and Section 4"},{"comment":"The treatment of unresolved galaxies and upper limits is unspecified. The text says that galaxies whose PSF-deconvolved fit fails are assigned upper limits from the non-deconvolved fit, and Fig. 6 plots upper limits, but the fit description says 'Based on all the valid IFU observations' without stating whether upper limits are included as censored data in the Kelly (2007) regression. Table 1 contains many entries with resolved=False. If these points are omitted, the low-luminosity end of the fit is biased toward resolved ENLRs; if they are included as detections, the slope is biased downward. Please specify the censoring scheme, report the fit with and without upper limits, and state how many galaxies required extrapolation of the Sersic profile beyond the data to reach R16.","section":"Section 3.4 and Section 4"},{"comment":"The agreement with the Dempsey & Zakamska (2018) model is presented as support, but the model has two free parameters (cloud mass mc and covering factor Omega) that are adjusted to match the data. The 'best' model in Fig. 8 is therefore a two-parameter fit to the same relation, not an a priori prediction, so the comparison is illustrative rather than a strong test. Please either provide a formal model comparison that accounts for the fitted parameters, or soften the Abstract/Conclusions wording to 'consistent with' rather than 'supports'.","section":"Section 5 and Figure 8"}],"minor_comments":[{"comment":"Several entries list log L[O III] uncertainties larger than 10 dex (e.g., 8483-12703, 9502-9101, 9893-6102); these values are unphysical and should be corrected or explained.","section":"Table 1"},{"comment":"The sentence 'Both the number of radio-loud AGN in our sample and their ENLR sizes These results suggest...' is grammatically incomplete; please rewrite.","section":"Section 5"},{"comment":"'1th fitting' and '2th fitting' should be '1st fitting' and '2nd fitting'.","section":"Figure 5 caption and Section 3.4"},{"comment":"The warning about cosmological dimming is placed in the MaNGA methods section, but the relevant application is to the quasar sample; please expand it where the combined fit is described and state the numerical correction used for each source.","section":"Section 3.3"},{"comment":"Please state explicitly how many galaxies are included in each fit (MaNGA-only and combined) and make clear in the figure/legend which points are detections and which are upper limits.","section":"Figure 6"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of MNRAS and the analysis is generally careful. The main load-bearing issue is the surface brightness threshold convention, which is easily testable by checking the Liu et al. (2013, 2014) definitions and re-running the fit. I would be satisfied if the revision states the convention explicitly, re-fits with a consistent threshold, and clarifies how unresolved/upper-limit galaxies enter the regression."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper deserves your attention, but check the surface-brightness convention before quoting the headline slope. The new piece is a large, uniform IFU sample: 152 MaNGA AGN with ENLRs identified by spatially resolved BPT diagrams. That is a real step forward I haven't seen done on this scale. The analysis is careful: PSF deconvolution with Sersic fits, Monte Carlo radius uncertainties, Balmer-decrement dust correction, and a Bayesian fit. The MaNGA-only slope of 0.49 ± 0.04 is a solid measurement, and the mock long-slit comparison nicely shows why IFU sizes beat slit-based ones. The paper is clearly written and the code is public.\n\nThe soft spots are in the combination with the literature quasars. Section 3.3 warns about cosmological dimming, but Section 4 never says whether the Liu et al. profiles were converted to rest frame before applying the 10^-16 threshold. If the quasar sizes are measured in the observed frame at z ~ 0.5-1, they are systematically too large, and the combined slope of 0.42 ± 0.02 could simply be the MaNGA slope pulled down by an inconsistent size definition. That is not a proven error - Liu et al. may already be quoting rest-frame radii - but the paper has to state it explicitly. This is the main fix I would ask for.\n\nThe model comparison in Fig. 8 is also weaker than it looks. The Dempsey & Zakamska (2018) curves are overlaid with two free parameters tuned to the data, and one of this paper's coauthors is a coauthor of that model. The agreement is illustrative, not a rigorous test. The conclusion that outflows and jets are not needed rests on nine radio-loud AGN and no direct outflow-size measurements; I would call that suggestive rather than demonstrated.\n\nNone of this kills the paper. If you work on AGN feedback, ENLR sizes, or IFU surveys, the sample and the low-luminosity constraint are genuinely useful. The threshold convention is a load-bearing ambiguity, and a serious referee can resolve it. I would send it out, with a request that the authors document the rest-frame correction or refit without it. Once that is settled, I would cite the paper for the MaNGA sample and the slope.","headline":"A genuinely useful IFU-based ENLR sample, but the headline 0.42 slope needs a clear statement that the quasar isophotal threshold is rest-frame.","tokens_in":27373,"tokens_out":5487,"would_cite":false,"duration_ms":51624,"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":"A single power law links the size of AGN-ionized gas to [O III] luminosity across four orders of magnitude.","keywords":["active galactic nuclei","extended narrow line region","size-luminosity relation","MaNGA survey","integral field spectroscopy","[O III] emission","photoionization","Seyfert galaxies"],"falsifier":"Re-measure the quasar sizes at the rest-frame equivalent of $10^{-16}\\,\\mathrm{erg\\,s^{-1}\\,cm^{-2}\\,arcsec^{-2}}$ by multiplying observed surface brightness by $(1+z)^4$, then re-fit the combined sample; if the slope moves by more than the quoted $0.02$ uncertainty, the claimed universal relation is an artifact of the threshold mismatch.","tokens_in":26242,"feed_emoji":"🔭","tokens_out":7575,"duration_ms":76763,"temperature":0.7,"pith_summary":"This paper sets out to measure how far AGN-ionized gas extends as a function of AGN luminosity, using a uniform integral-field sample rather than mixed long-slit data. It reports that the size of the extended narrow line region at a fixed surface brightness follows $\\log(R/\\mathrm{pc}) = (0.42\\pm0.02)\\log(L_{\\mathrm{[O\\,III]}}/\\mathrm{erg\\,s^{-1}}) - (13.97\\pm0.95)$ across four orders of magnitude in [O III] luminosity, from nearby Seyferts to luminous quasars. A sympathetic reader would care because the slope discriminates between models: it is close to the 0.5 of simple photoionization and to the 0.45 predicted by a cloud-population model, and it implies radiation alone can push ionized gas to kiloparsec scales without outflows or jets. The result also suggests that previous slope disagreements came largely from small samples and from slit-based size definitions that underestimate the true ENLR extent.","feed_headline":"AGN light sets ionized-gas size across four decades","feed_subtitle":"The same 0.42-slope relation links nearby Seyferts to luminous quasars, favoring pure photoionization.","key_machinery":"The argument runs on three linked tools: spatially resolved BPT classification using the Kewley et al. (2001) boundary to select AGN-dominated spaxels; summing the dust-corrected [O III] flux of those spaxels as the AGN luminosity; and fitting the [O III] surface-brightness profile with a Sersic law, with and without convolution by the MaNGA PSF, to read off R16. The threshold-based radius is the same quantity the quasar studies used, which is what makes the four-dex combination possible.","core_discovery":"The central claim is that the ENLR size, defined as the radius at the $10^{-16}\\,\\mathrm{erg\\,s^{-1}\\,cm^{-2}\\,arcsec^{-2}}$ [O III] surface-brightness isophote, is a single power-law function of AGN [O III] luminosity: $\\log(R/\\mathrm{pc}) = (0.42\\pm0.02)\\log(L_{\\mathrm{[O\\,III]}}/\\mathrm{erg\\,s^{-1}}) - (13.97\\pm0.95)$. The relationship holds over four dex when the MaNGA Seyferts are combined with IFU quasar measurements, and the slope is insensitive to whether radio-loud AGN or galaxies with detected outflows are present. This supports the picture of an ENLR as a population of photoionized clouds in pressure equilibrium with the radiation field, rather than gas mechanically transported by jets.","pith_inferences":["If the surface-brightness threshold was left in observed frame, the quasar R16 values are measured at a fainter intrinsic brightness than the MaNGA values; correcting with $(1+z)^4$ and refitting would test whether the 0.42 slope survives.","The same method applied to JWST/NIRSpec IFU observations at $z\\sim1\\text{--}3$ could test whether the relation holds at higher redshift or turns over.","Combining ENLR sizes with molecular-gas maps would test whether the $n_c\\propto r^{-2}$ cloud distribution assumed by the model is physically present."],"forward_implications":["The 0.42 slope implies ENLR sizes scale roughly as $L^{0.4}$, a scaling that standard photoionization and the Dempsey & Zakamska cloud-population model can both accommodate.","The MaNGA-only slope of $0.49\\pm0.04$ is consistent with the combined slope, suggesting the same physical scaling connects nearby Seyferts and luminous quasars.","Radio-loud AGN, nine of the 152 galaxies, lie on the same relation, so jets are not the dominant driver of ENLR size in this luminosity range.","Because PSF-smearing corrections change measured sizes by up to about 50 percent for marginally resolved sources, PSF-corrected sizes are essential when comparing IFU surveys.","If the relation is universal, the spatial extent of ionized gas alone can serve as a rough luminosity indicator for AGN whose broad lines are hidden."],"supporting_citations":[{"why":"Defines the MaNGA survey whose MPL-8 data supply the 152 AGN sample.","marker":"Bundy et al. (2015)"},{"why":"Provides the spaxel-based AGN selection procedure adapted in this work.","marker":"Rembold et al. (2017)"},{"why":"Supplies the maximum-starburst boundary (Ke01) used to classify AGN-dominated spaxels.","marker":"Kewley et al. (2001)"},{"why":"Provides IFU [O III] surface-brightness profiles of luminous type-II quasars that extend the sample to high luminosity.","marker":"Liu et al. (2013)"},{"why":"Adds type-I quasar IFU measurements that complete the four-dex luminosity range.","marker":"Liu et al. (2014)"},{"why":"The cloud-population photoionization model whose predicted slope is compared with and fits the observed relation.","marker":"Dempsey & Zakamska (2018)"},{"why":"The Bayesian linear-regression method used to derive the best-fit slope and intercept.","marker":"Kelly (2007)"}],"fun_headline_variants":["ENLR size scales with AGN luminosity, slope 0.42","One power law links AGN sizes from Seyferts to quasars","Photoionization sets AGN ionized-gas size, no jets needed","MaNGA AGN: extended narrow-line region size over 4 dex","AGN radiation alone explains narrow-line region sizes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fit assumes that measuring sizes at the same observed-frame surface brightness of $10^{-16}\\,\\mathrm{erg\\,s^{-1}\\,cm^{-2}\\,arcsec^{-2}}$ is equivalent for the nearby MaNGA galaxies and the higher-redshift quasars; if the threshold was not converted to rest frame, the quasar sizes sit at a different intrinsic brightness and the combined slope could be biased.","fun_headline_variants_meta":{"raw":{"variants":["ENLR size scales with AGN luminosity, slope 0.42","One power law links AGN sizes from Seyferts to quasars","Photoionization sets AGN ionized-gas size, no jets needed","MaNGA AGN: extended narrow-line region size over 4 dex","AGN radiation alone explains narrow-line region sizes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000711,"raw_usage":{"total_tokens":3190,"prompt_tokens":924,"completion_tokens":2266,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":540,"completion_tokens_details":{"reasoning_tokens":2174}},"tokens_in":540,"tokens_out":2266,"duration_ms":17277,"temperature":1.0,"reasoning_tokens":2174,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:31:01.047354+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure the quasar sizes at the rest-frame equivalent of $10^{-16}\\,\\mathrm{erg\\,s^{-1}\\,cm^{-2}\\,arcsec^{-2}}$ by multiplying observed surface brightness by $(1+z)^4$, then re-fit the combined sample; if the slope moves by more than the quoted $0.02$ uncertainty, the claimed universal relation is an artifact of the threshold mismatch.","supporting_citations":[],"review_version":1}