{"id":"1a938378-ba05-4ded-a003-a053f8851406","arxiv_id":"2501.00806","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Two X-ray lamps at different heights above an AGN disk produce UV/optical delays at most 20 percent different from a single lamp, and simulated ICCFs are indistinguishable.","lead":"This paper simulates how an accretion disk lit by two X-ray lamps at different heights above the black hole would appear, and compares it with the usual single-lamp model. It finds that the two setups produce nearly identical UV/optical delays, so current reverberation mapping cannot easily reveal the vertical size of the corona.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero-lag, perfectly coherent two lamps may understate sensitivity of ICCF to vertical corona extent.","rationale":"The reader's identified weakest assumption is precisely the zero-lag, identical-lightcurve approximation for the two lamps, which is the same concern I find most load-bearing. The paper's two principal evidence streams both depend on this assumption: the transfer-function delay curves (which assume simultaneous pulses) and the ICCF simulations (which use the same dense curve for both coronae). Relaxing it could plausibly make the two-lamp model more distinguishable, either through asymmetric response kernels or through differences in the cross-correlation function shape. The concern is not that the authors are wrong within their adopted model, but that the model excludes a physically relevant degree of freedom (internal lags/coherence) without a sensitivity test. The reader's CONDITIONAL verdict already anticipates that such gaps should be closed, so I do not recommend changing the verdict; the paper should be accepted only if the authors either justify the zero-lag assumption quantitatively or demonstrate that the conclusion is robust to plausible lags and partial coherence. My concrete test would settle this directly.","tokens_in":15809,"tokens_out":12647,"duration_ms":114116,"concrete_test":"Repeat the ICCF simulations of Section 3.3 with the same two-lamp parameters (h1=5, h2=100 Rg, L1/L2=3, M=1e8 Msun, Eddington=1) but introduce a relative time delay tau = (h2-h1)/c = 0.55 days between the two lamps' driving light curves, and also a partially coherent case (correlation coefficient rho=0.5). Run N=100 realizations, compute the mean and dispersion of the ICCF peak lags, and perform a two-sample significance test (e.g., Welch's t-test) comparing the single-lamp and two-lamp peak-lag distributions. If the distributions separate by more than 3 sigma, the zero-lag assumption is load-bearing and the ICCF null result does not generalize. Additionally, recompute the transfer function for the two lamps with a delay delta=0.55 days between the two pulses and re-derive Table 1's maximum deviation at 1000 A to check whether the 20% ceiling still holds.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that UV/optical reverberation mapping is insensitive to vertical corona extension rests on treating the extended corona as two lamps that share exactly the same variability time series with no intrinsic delay (Section 2.3: 'we usually use the same dense curve, without any intrinsic delay between the coronal points'). This is a measure-zero subset of possible correlations for a real continuous corona, whose top and bottom will experience light-crossing delays (for h2-h1 = 95 Rg and M = 1e8 Msun, tau = 95*GM/c^3 = 0.5 days) and may have partial coherence. Such lags or partial coherence could introduce asymmetries in the response function and in the resulting ICCF, potentially breaking the degeneracy that the paper reports. The 20% delay-curve ceiling and the ICCF null result are both computed under this zero-lag assumption, and the paper does not test how the conclusion changes when it is relaxed. The paper explicitly acknowledges the assumption but offers no justification or sensitivity study, so the generality of the conclusion 'not very sensitive' is unsupported for realistic coronae.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper tests whether optical/UV reverberation mapping can distinguish a vertically extended X-ray corona from a single point-like lamppost. The extended corona is modeled as two point sources on the symmetry axis at heights h1=5 Rg and h2=100 Rg, with variable luminosity ratio, for a 1e8 solar mass black hole at Eddington ratio 1 and inclination 30 degrees. Using a Shakura-Sunyaev disk with perfect thermalization of incident flux, the authors compute transfer functions and wavelength-dependent delays (Sections 2–3), finding that after matching the single-lamp height at the longest wavelength, the maximum delay difference at 1000 Å is about 20% for a luminosity ratio of ~3 (Table 1). They further simulate light curves with a Timmer-Koenig power spectral density and compute ICCFs between 1000 Å and 7923 Å for 10 realizations; the single-lamp and two-lamp ICCFs overlap within the dispersion (Sections 3.3–3.4). The paper concludes that UV/optical reverberation mapping is not very sensitive to vertical corona extension and that the effective lamp position is a luminosity-weighted average height (Eq. 11).","tokens_in":16062,"tokens_out":7938,"duration_ms":68490,"significance":"If correct, the result is a valuable caution for the AGN reverberation-mapping community: corona heights fitted from continuum delays should be interpreted as luminosity-weighted averages rather than as direct geometric measurements of the vertical extent. The paper is a well-defined, falsifiable test: it uses standard thin-disk reprocessing, transfer functions, and identical driving light curves for the single- and two-lamp cases, which avoids the circularity of fitting to data. The central 20% deviation and the ICCF null result are concrete predictions for the chosen parameter set. The main limitations are the idealized two-lamp representation with perfectly coherent, zero-lag variability, and the lack of a formal statistical test on the ICCF comparison; these do not invalidate the specific calculations but do bound the generality of the conclusion.","major_comments":[{"comment":"The claim that the ICCF profiles of the two-lamp and single-lamp models are 'statistically indistinguishable' is not supported by a formal test. The paper compares mean ICCFs and dispersions over 10 realizations but does not report a p-value, confidence interval, or any null-hypothesis test for the peak-delay differences of 1.0 day (L1/L2=1) and 1.0 day (L1/L2=3) in Table 2. Because the realization-to-realization scatter is visibly large (Figure 6), a quantitative test is required before concluding that the models cannot be distinguished. Please add a permutation or bootstrap test that quantifies the probability of observing the measured mean shift under the null hypothesis of identical ICCF distributions.","section":"Sections 3.3–3.4, Table 2"},{"comment":"The assumption that both coronal components use 'the same dense curve, without any intrinsic delay between the coronal points' is a measure-zero subset of possible correlations in a real vertically extended corona. For the adopted parameters, the light-crossing time between h1=5 Rg and h2=100 Rg is about 0.5 days (for M=1e8 Msun), comparable to the sampling interval and to the delays being measured. If the upper and lower components are mutually delayed or only partially coherent, the two-lamp response function becomes asymmetric, potentially breaking the degeneracy with a single lamppost that the paper reports. The paper acknowledges this limitation but does not explore it. Please compute at least one test case with a fixed intrinsic delay between the two lamps (e.g., 0.5 days) and one with a coherence factor less than unity, and report how the 20% ceiling and the ICCF result change. Without this, the general conclusion that UV/optical reverberation mapping is insensitive to vertical extension is not established.","section":"Section 2.3"},{"comment":"The central 20% result is presented as the maximum deviation of the wavelength-dependent delay curve, but Table 1 and Section 3.2 evaluate the deviation at a single wavelength, 1000 Å. The paper does not show that the deviation is maximal at 1000 Å over the range 1000–10000 Å, and indeed the text states the departure is measured 'at the shortest wavelengths,' leaving it possible that the deviation grows further below 1000 Å. Please present the deviation as a function of wavelength over the full computed range, or explicitly state that the 20% applies to 1000 Å only and adjust the abstract and conclusions accordingly.","section":"Abstract, Table 1, Figure 5"}],"minor_comments":[{"comment":"The notation for the lamp luminosities is inconsistent: Eq. (1) uses L12 and L21 for the two sources, while Eqs. (3)–(4) use L12 and L22, and the text says 'L12 and L21 are luminosities associated with the two sources.' Please use L1 and L2 throughout to avoid confusion.","section":"Eqs. (1)–(4), Section 2.3"},{"comment":"The key words list 'giant planet formation – κ-mechanism – stability of gas spheres' appears to be a copy-paste error from an unrelated paper; it should be replaced with keywords appropriate to AGN accretion disk reverberation mapping.","section":"Key words line"},{"comment":"The text states the two-lamp response function 'can be approximated by summing the red and blue single-corona response functions,' but in the figure the two single-corona response functions are blue (h=5 Rg) and green (h=100 Rg). This should read 'blue and green.'","section":"Figure 3 and Section 3.1"},{"comment":"The paper states that with 10 light curves 'the error of the mean would be a factor of sqrt(N) lower,' but it does not show the standard error of the mean in Figure 7; adding error bars to the mean ICCF curves would make the comparison more transparent.","section":"Section 3.4"},{"comment":"The paper uses 'lamppost' and 'lampost' inconsistently; please standardize spelling.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid, focused contribution that fits A&A's scope. The central calculations are straightforward and internally consistent, and the use of identical driving light curves for both models is a strength. The requested sensitivity tests (intrinsic lag/coherence, full-wavelength deviation, formal ICCF statistics) are essential before the general insensitivity claim can be accepted. The manuscript would also benefit from a brief discussion of how inclination (fixed at 30 deg) affects the results, and from making the simulation code available for reproducibility."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know before you read it. First, this is the first systematic test of whether a vertically extended corona, modeled as two discrete lamps, can be distinguished from a single lamppost using UV/optical continuum delays and ICCFs. The answer is mostly no: the wavelength-dependent delay curve deviates by at most about 20% from the best-matched single lamp, and the ICCF is statistically indistinguishable in 10 realizations. Second, the conclusion is conditional on an assumption the paper acknowledges but does not test: the two lamps share exactly the same variability time series, with no intrinsic lag or partial coherence. That assumption is doing real work.\n\nWhat is genuinely new: prior extended-corona studies targeted X-ray spectra and timing. This paper targets the optical/UV reprocessing band, computes transfer functions for a two-lamp geometry, fits the single-lamp height to the longest wavelength, and quantifies the deviation across wavelength and luminosity ratio. The luminosity-weighted effective height (Eq. 11) is a clean interpretive result that should help observers who fit continuum delays. The numerical setup is also fair in one respect: the same driving light curve was used for both models, so the comparison isolates geometry from stochastic realization noise.\n\nWhere the soft spots are. The zero-lag, coherent-lamps assumption is the main one. For a real corona, the upper and lower parts see a light-crossing time of order 0.5 days for 95 Rg at 1e8 M_sun, not negligible compared to the delays being measured. If the vertical parts vary with a lag or coherence less than unity, the combined response can become asymmetric in ways the two-lamp setup never explores. The paper says 'we usually use the same dense curve' but does not justify why that is the relevant physical case. So the strong statement 'not very sensitive to the vertical extension' holds for perfectly synchronized point sources; for a continuous, partially coherent corona it is an upper bound on sensitivity, not a proven result.\n\nThere are also reproducibility gaps. No code or data are shipped, the Timmer-Koenig PSD parameters (three slopes, two breaks) are not specified, the 20% number has no numerical uncertainty, and the ICCF null result rests on 10 realizations with no formal test. These are fixable, not fatal. The authors are honest that the high-mass, low-Eddington regime where differences become large is outside the validity of their own thin-disk, no-GR assumptions.\n\nBottom line: this is a useful, clearly scoped simulation that deserves a serious referee. The central result is plausible, but the zero-lag assumption needs to be relaxed or defended before the conclusion is generalized. I would bring it to the reading group, and I would probably cite it as a cautionary boundary when interpreting continuum RM heights.","headline":"A useful, clearly scoped simulation showing that two synchronous corona lamps mimic a single lamppost in UV/optical delays to within ~20%, but the zero-lag coherent-lamp assumption leaves the broad conclusion under-supported.","tokens_in":16591,"tokens_out":3159,"would_cite":true,"duration_ms":30361,"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 paper claims that UV/optical reverberation mapping cannot reliably distinguish a vertically extended X-ray corona from a point-like lamppost: the two-lamp delay curve differs from the best single lamp by at most about 20%, and…","keywords":["active galactic nuclei","accretion disks","corona","reverberation mapping","lamppost model","time delays","X-ray reprocessing","transfer functions"],"falsifier":"Measure the wavelength-dependent continuum delay curve of an AGN with $M_{\\rm BH} \\approx 10^8\\,M_\\odot$ and Eddington ratio near 1 using sampling better than 0.1 days over at least 200 days; if the delay curve cannot be matched by any single lamppost height within 20% at 1000 Å, or if the ICCF of the two-lamp model is systematically distinguishable from the single-lamp model, the claimed insensitivity fails. Alternatively, a high-mass, low-Eddington source should show the predicted sharp break in short-wavelength delays if the extended-corona interpretation is correct.","tokens_in":15590,"feed_emoji":"🔭","tokens_out":5925,"duration_ms":54812,"temperature":0.7,"pith_summary":"The paper asks whether optical/UV reverberation mapping—the standard technique for measuring where X-ray irradiation of an accretion disk comes from—can detect that the corona is vertically extended rather than point-like. To test this, the authors replace the usual single lamppost with two point sources at heights of 5 and 100 gravitational radii irradiating a standard thin disk, and compare the predicted wavelength-dependent time delays and cross-correlation functions with the best-fit single lamp. They find that the two-lamp delay curve never deviates by more than about 20% from a single lamp, with the largest deviation when the lower lamp is about three times brighter than the upper, and that simulated light curves with 0.1-day sampling over 200 days produce statistically indistinguishable ICCF profiles. They conclude that UV/optical reverberation mapping is not very sensitive to the vertical extension of the corona, so observer-fitted corona heights represent a luminosity-weighted average position rather than a direct measurement of the vertical extent.","feed_headline":"Stretched coronas leave almost no trace in UV/optical echoes","feed_subtitle":"Two-lamp corona mimics a single lamppost within 20 percent; fitted heights are luminosity-weighted averages.","key_machinery":"The central object is the wavelength-dependent transfer (response) function of a thin accretion disk irradiated by two point sources on the symmetry axis. The paper computes the delay from each lamp to each disk element plus the path to the observer, assumes perfect thermalization of the incident X-ray flux, and derives the reprocessed temperature and monochromatic luminosity as a function of wavelength. The argument that a two-lamp setup mimics a single lamp rests on the identity that at large radii the ratio of incident to dissipative flux is proportional to the luminosity-weighted sum $\\sum_i L_i H_i$, so the effective height is the luminosity-weighted average position of the emitting components.","core_discovery":"The central claim is that the UV/optical continuum reverberation signal is produced by reprocessing at radii large compared with the corona height, so the observable delays depend on the luminosity-weighted average height $h = (h_1 L_1 + h_2 L_2)/(L_1 + L_2)$ rather than on the detailed vertical distribution of the X-ray source. For a $10^8\\,M_\\odot$ black hole at Eddington ratio 1 with lamps at 5 and 100 $R_g$, the wavelength-dependent delay of the two-lamp model differs from the best-matching single lamppost by at most about 20%, occurring near a luminosity ratio of 3, and the ICCFs of simulated 0.1-day-sampled light curves show no statistically measurable difference between the two geometries. The authors therefore conclude that UV/optical reverberation mapping cannot reliably reveal a vertically extended corona, and that fitted heights from such data should be read as effective, luminosity-weighted positions.","pith_inferences":["Because the paper assumes the two lamps share exactly the same variability curve with no intrinsic delay, a real continuous corona whose subregions fluctuate with mutual lags or partial coherence could produce asymmetries larger than the 20% ceiling; testing this would require multi-zone variability models rather than two synchronous lamps.","The luminosity-weighted-average result suggests that combining X-ray and optical/UV reverberation—probing inner and outer disk regions respectively—might recover some vertical information even though each band alone cannot, because the effective height should be wavelength-dependent if the corona is extended.","A testable extension of the paper's logic is to search existing multi-band continuum reverberation data for a systematic trend in the best-fit single-lamp height with wavelength; such a trend would be evidence for vertical extension beyond the point-source approximation.","The paper's prediction for high-mass, low-Eddington sources—a sharp flattening of short-wavelength delays—could be checked with future ultraviolet monitoring campaigns, but the prediction currently rests on Newtonian optics and a standard disk extending to the ISCO, which may not hold in that regime."],"forward_implications":["Continuum reverberation campaigns that fit a single corona height are measuring a luminosity-weighted mean height, not the true vertical extent; a corona spanning 5 to 100 $R_g$ with comparable luminosities will look like a lamp near 52 $R_g$.","Distinguishing an extended corona from a point source requires wide wavelength coverage from about 1000 to 10000 Å free of broad-line-region contamination, time-delay errors below 20%, and sampling much denser than the 0.1-day cadence used here.","The ICCF peak is statistically shifted to longer delays for the two-lamp model (for example 5.69 days versus 4.69 days at equal luminosities), but the shift is smaller than the dispersion among realizations, so a single monitoring campaign cannot exploit it.","For higher black hole mass ($10^9\\,M_\\odot$) and lower Eddington ratio (0.01), the two-lamp delay curve bends sharply away from the standard $\\lambda^{4/3}$ trend at short wavelengths, offering the best chance to see the extension, though relativistic modeling would be required.","When one lamp strongly dominates, the two-lamp model is indistinguishable from a single lamp at the dominant height, so the detectable signature is confined to a narrow range of luminosity ratios, roughly 2 to 4."],"supporting_citations":[{"why":"Supplies the standard thin-disk model whose dissipative flux and temperature the irradiation terms are added to.","marker":"Shakura & Sunyaev (1973)"},{"why":"Provides the disk-grid and delay-computation method used here for single-lamp delay profiles, which the two-lamp comparison extends.","marker":"Jaiswal et al. (2023)"},{"why":"Establishes the wavelength-dependent delay calculations and the pulse-width limit that the transfer-function computation follows.","marker":"Kammoun et al. (2021a)"},{"why":"Provides the perfect-thermalization treatment for incident X-ray flux used to compute the reprocessed disk temperature and emission.","marker":"Kammoun et al. (2019)"},{"why":"Supplies the algorithm for generating the synthetic coronal light curves that drive the ICCF simulations.","marker":"Timmer & Koenig (1995)"},{"why":"Gives representative corona heights derived from data fitting that the paper contrasts with X-ray inferred heights and reinterprets as luminosity-weighted averages.","marker":"Kammoun et al. (2021b)"},{"why":"Identified as the advanced modeling route needed to test the extended-corona predictions in the high-mass, low-Eddington regime.","marker":"Langis et al. (2024)"}],"fun_headline_variants":["Extended corona leaves no trace in UV/optical echoes","UV/optical reverberation blind to corona vertical extent","Two-lamp corona mimics single lamppost in UV/optical","Corona shape hidden in UV/optical disk echoes","UV/optical delays measure only average corona height"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results assume the two coronal components follow exactly the same variability pattern with no intrinsic delay between them; a real extended corona whose subregions fluctuate with lags or partial coherence could produce stronger or asymmetric signatures that the 20% ceiling and ICCF null result would miss.","fun_headline_variants_meta":{"raw":{"variants":["Extended corona leaves no trace in UV/optical echoes","UV/optical reverberation blind to corona vertical extent","Two-lamp corona mimics single lamppost in UV/optical","Corona shape hidden in UV/optical disk echoes","UV/optical delays measure only average corona height"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000269,"raw_usage":{"total_tokens":1688,"prompt_tokens":1077,"completion_tokens":611,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":693,"completion_tokens_details":{"reasoning_tokens":533}},"tokens_in":693,"tokens_out":611,"duration_ms":5431,"temperature":1.0,"reasoning_tokens":533,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:42:04.386349+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the wavelength-dependent continuum delay curve of an AGN with $M_{\\rm BH} \\approx 10^8\\,M_\\odot$ and Eddington ratio near 1 using sampling better than 0.1 days over at least 200 days; if the delay curve cannot be matched by any single lamppost height within 20% at 1000 Å, or if the ICCF of the two-lamp model is systematically distinguishable from the single-lamp model, the claimed insensitivity fails. Alternatively, a high-mass, low-Eddington source should show the predicted sharp break in short-wavelength delays if the extended-corona interpretation is correct.","supporting_citations":[{"cited_title":"K., Prince, R., Panda, S., & Czerny, B","cited_arxiv_id":null,"evidence_quote":"Provides the disk-grid and delay-computation method used here for single-lamp delay profiles, which the two-lamp comparison extends."},{"cited_title":"S., Papadakis, I","cited_arxiv_id":null,"evidence_quote":"Provides the perfect-thermalization treatment for incident X-ray flux used to compute the reprocessed disk temperature and emission."},{"cited_title":"& Koenig, M","cited_arxiv_id":null,"evidence_quote":"Supplies the algorithm for generating the synthetic coronal light curves that drive the ICCF simulations."}],"review_version":1}