{"id":"4f0d4fe1-2fc8-49c1-8259-5e6530be92b7","arxiv_id":"2505.15881","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Stacked imaging of 853 high-redshift galaxies shows extended Lyman-alpha halos around all subsamples, including central absorbers, with scale lengths of 10-20 kpc that depend only weakly on galaxy properties.","lead":"This paper stacks images of 853 galaxies seen when the universe was about 2.6 billion years old and finds that nearly all of them glow in Lyman-alpha light far beyond their stars. The glow appears even around galaxies whose centers absorb Lyman-alpha, suggesting halos are a generic feature of star-forming galaxies.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central absorption scale-length claim is vulnerable to PSF/model degeneracy; recovery tests are needed.","rationale":"The reader identified the empirical PSF and background subtraction as the fragile premise; I agree on the PSF but locate the specific failure mode in the identifiability of the negative exponential component used to claim spatial correspondence. The net-emission result is less vulnerable because aperture photometry is PSF-conserving and the background annulus (8–11 arcsec) makes the measurement differential and conservative. The central absorption scale claim, however, depends on a model decomposition that has not been validated against known inputs. The recommended verdict remains CONDITIONAL, matching the reader's verdict, with an added condition: the authors should demonstrate via injection-recovery that r0,gal is recovered without bias for negative-amplitude galaxy components and that the result is insensitive to reasonable PSF variations. If the recovery test fails, the second half of the abstract's central claim should be softened. This is not a rejection because the primary net-emission result and the overall data products appear sound, and the stated uncertainty caveats in Sec. 5.2 and 5.3 already acknowledge sensitivity to stacking choices.","tokens_in":34298,"tokens_out":8912,"duration_ms":87773,"concrete_test":"Run an end-to-end recovery simulation. Construct a noiseless Lyα image from a negative exponential galaxy component (r0=1.8 kpc, amplitude matched to KBSS Abs) plus a positive exponential halo (r0=11 kpc), convolved with the final empirical PSF, with the same background subtraction as Sec. 3.3. Add bootstrap noise and fit with the same MCMC forward model. Repeat using (i) the average PSF, (ii) each of the 18 individual field PSFs from Fig. 15, and (iii) a Moffat PSF truncated at 5 arcsec. Also repeat with widened priors (e.g., r0,gal<10 kpc). If the recovered r0,gal varies by more than 0.5 kpc across PSF choices, or if the posterior is prior-dominated for the Abs/All stacks, the 'spatial scale of central Lyα matches continuum' claim is not robust. Report the posterior vs. prior for r0,gal in all cases.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 4.2's forward model fits a negative exponential 'galaxy' component for the KBSS Abs and All stacks, and Sec. 5.2/Fig. 11 use the recovered r0,gal to claim that the spatial scale of central Lyα absorption matches the continuum. This is load-bearing for the abstract. The inference requires that the empirical stellar PSF (Appendix A) accurately represents the light spread of an extended negative surface-brightness feature, and that a negative compact exponential is identifiable against a positive halo plus PSF wings. The model imposes hard priors r0,gal<5kpc and r0,halo>5kpc (footnote 21), and for KBSS All r0,gal is only a 2σ upper limit (<2.7kpc), so the 'match' is not actually constrained for the full KBSS stack. For KBSS Abs, the quoted 1.7±0.3 kpc is precise, but the fixed stacking procedure and the authors' own statement (Sec. 5.3) that quoted uncertainties are lower bounds mean this precision may not reflect systematic PSF/stacking variation. A mismatch between the true extended-source PSF and the stellar PSF at 1–5 arcsec would be absorbed by the negative exponential, biasing its scale length toward the continuum scale. The claimed spatial correspondence could then be an artifact of the assumed PSF and priors rather than a physical measurement.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the Keck/LRIS narrowband Lyα imaging of the KBSS-Lyα survey for the first time and uses stacked imaging of 734 Lyα-selected and 119 continuum-selected galaxies at 250 kpc; that the forward-modeled central Lyα component has a scale length consistent with the continuum scale length, including for the absorbing subsample; and that the halo scale length shows only weak dependence on M_UV and EW_Lyα, possibly in the sense of larger halos for fainter, high-EW galaxies. The paper also compares the results with earlier KBSS, MUSE, and Subaru measurements.","tokens_in":34491,"tokens_out":6134,"duration_ms":59865,"significance":"If the central claims hold, the paper would establish that Lyα halos are a generic property of star-forming galaxies at z~2-3, extending even to galaxies whose small-aperture spectra show net Lyα absorption, and it would provide the first quantitative comparison of the spatial scale of central Lyα absorption with the stellar continuum. The paper has notable strengths: it presents a large, homogeneous imaging data set; it cross-checks the halo measurements with two independent fitting methods; it provides bootstrap uncertainties; and it explicitly acknowledges that the fixed stacking methodology makes reported uncertainties lower bounds on the total systematic uncertainty (Sec. 5.3). The near-equality of f_gal^Lyα with the ratio of small-aperture to 50 kpc EW_Lyα is correctly identified by the authors as a consistency check rather than independent evidence (Sec. 5.1). The most fragile load-bearing claim is the scale match of the central Lyα absorption component, which relies on the recovery of a negative compact exponential against the positive halo and the PSF wings without a demonstrated recovery test.","major_comments":[{"comment":"The abstract states that the spatial scale of central Lyα 'approximately matches' that of the continuum emission, and Sec. 5.2 extends this to the KBSS All and Abs stacks where the central Lyα component is negative. This claim is not equally supported for all stacks: for KBSS All, Table 3 reports only a 2σ upper limit r0,gal<2.7 kpc, so the scale of the absorbing component is not actually measured for the full KBSS stack. For KBSS Abs, the quoted r0,gal=1.7±0.3 kpc is precise, but the paper presents no injection-recovery test demonstrating that a negative compact exponential can be recovered when convolved with the empirical PSF and superposed on a positive halo. Without such a test, the agreement with r0,cont=1.8±0.1 kpc may reflect the model priors and the PSF-wing shape rather than a physical co-location of absorbing gas and stars. I request a recovery test using synthetic stacks with known negative galaxy components and positive halos, fit through the same pipeline, and a revised abstract that restricts the scale-match claim to the stacks for which it is actually constrained.","section":"Sec. 4.2, Table 3, Fig. 11, and Abstract"},{"comment":"The forward-modeling results depend on the assumption that the empirical stellar PSF, after iterative smoothing to a common 'broadest PSF' and with the outer region (θ>10'') replaced by a Moffat model, accurately describes the light spread of the extended galaxy-plus-halo emission at all fitted radii. A mismatch between the stellar PSF and the true extended-source PSF at intermediate radii (≈1-5 arcsec) would be partially absorbed by the central exponential terms, biasing r0,gal toward the continuum scale and also affecting r0,halo. The paper currently provides no quantitative test of this sensitivity. I recommend repeating the fits with alternative PSF constructions (e.g., unsmoothed field PSFs, pure Moffat PSFs, different transition radii, and PSFs with artificially modified wing amplitude) and reporting the resulting scatter in r0,gal and r0,halo as a systematic uncertainty. This is a necessary check for the specific claim of co-located central absorption and continuum emission.","section":"Appendix A and Sec. 4.2"},{"comment":"The claimed weak trends of halo scale length with M_UV and EW_Lyα are not robustly established by the present analysis. In Table 3, the direct exponential fits give r0,dir(LoEW)=18.5±3.4 kpc versus r0,dir(HiEW)=20.2±2.5 kpc, and r0,dir(Faint)=21.3±5.0 versus r0,dir(Bright)=21.2±2.8 kpc, i.e., no significant difference; the apparent trends come from the forward-modeled r0,halo values with asymmetric bootstrap uncertainties. Because the stacking method is fixed, these uncertainties do not include the effect of different sigma-clipping thresholds, weighting, or background corrections, which the authors themselves note in Sec. 5.3 would increase the inferred variation particularly for the KBSS All and Abs stacks. The paper should either marginalize over a range of stacking choices and report the resulting uncertainties on the trend, or explicitly present the trend as not significant pending such an analysis. The current wording in Summary item 6 ('we find a weakly negative relationship... as well as a weak positive correlation') overstates the evidence given the acknowledged systematics.","section":"Sec. 5.3, Table 3, Fig. 13, and Summary item 6"}],"minor_comments":[{"comment":"The phrase 'other recent students of Lyαhalos' should read 'other recent studies of Lyα halos'.","section":"Sec. 4.2"},{"comment":"In the footnote describing the S. Kikuta et al. (2023) models, 'our on work' should read 'our own work'.","section":"Sec. 6.3"},{"comment":"The word 'correponding' on the NB filter description should be 'corresponding'.","section":"Sec. 2.1"},{"comment":"The sentence containing 'all of the the observed profiles' has a duplicated 'the' and should be corrected.","section":"Sec. 6.1.2"},{"comment":"The caption of Figure 5 describes the continuum profiles as 'as in Fig. 5', but the intended cross-reference is likely to Fig. 4.","section":"Fig. 5 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational contribution with a large new data set and a generally careful treatment of stacking and PSF effects. My main concern is the central absorption scale-length claim, which currently rests on a forward-modeling recovery that is not demonstrated. Adding injection-recovery tests and PSF-alternative systematic estimates would substantially strengthen the paper. I would also ask the authors to make the weak-trend language more cautious unless they can quantify stacking-method systematics. These issues are addressable within the scope of the manuscript, so I do not recommend rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Roughly: the paper does what it says, and it is more honest than most. The new thing is the first release of the KBSS-Lyα imaging and the forward-modeling fits to stacks with net central Lyα absorption. The two fitting methods agree, bootstrap errors are provided, and the authors are explicit that their quoted errors are lower bounds given a fixed stacking procedure. That is the right way to present this kind of measurement, and it earns the reader's trust.\n\nThe central qualitative result—all stacks, including central absorbers, show net Lyα emission at R>50 kpc—looks solid. The direct and forward-modeled fits are consistent, and the comparison to Steidel+11 and MUSE is reasonable. The curve-of-growth section is a nice addition; it makes the aperture dependence of EW explicit. The near-tautology between f_gal^Lyα and the EW ratio is explicitly labeled a consistency check, not evidence, so there is no circularity problem.\n\nThe soft spots are real but not fatal. First, the claim that the spatial scale of central Lyα absorption matches the continuum (r0,gal ≈ r0,cont) is load-bearing for the abstract. For the KBSS All stack, r0,gal is only a 2σ upper limit (<2.7 kpc), so the match is not actually constrained. For KBSS Abs, the quoted 1.7±0.3 kpc is precise, but that precision comes from a fixed stacking methodology, and the authors admit the uncertainties are lower bounds. The stress-test concern is legitimate: if the true extended-source PSF differs from the stellar PSF at 1–5 arcsec, a negative exponential component would soak up the difference and its scale length would be pulled toward the continuum scale. The hard priors (r0,gal<5 kpc, r0,halo>5 kpc) make it worse. The paper should include recovery tests—inject a known negative exponential into stacks with a mismatched PSF and show the fit recovers the input scale length. Without that, the spatial correspondence claim is plausible but not demonstrated.\n\nSecond, the weak trends with M_UV and EW are exactly the kind of result that could flip with different stacking choices. The authors hedge appropriately, but the fixed-stacking caveat means those trends are provisional. Data release would help a lot; the paper currently does not provide the imaging data or catalogs, which makes independent replication difficult.\n\nWho is this for? Anyone working on Lyα halos or the CGM at z~2–3. It deserves a serious referee, and the referee should ask for recovery tests and data release. I would accept it for review as is, with those items as likely conditions for publication. The paper is a genuine step forward; the caveats are manageable.","headline":"A solid, honest data paper with one genuinely new capability (fitting central Lyα absorbers), but the headline spatial-match claim needs a robustness check with recovery tests before it can be taken at face value.","tokens_in":35165,"tokens_out":2553,"would_cite":true,"duration_ms":22643,"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":"Galaxies whose cores absorb Lyman-alpha still emit it at 50 kpc and beyond.","keywords":["Lyman-alpha emission","circumgalactic medium","galaxy stacks","point-spread function","high-redshift galaxies","narrowband imaging","equivalent width","halo scale length"],"falsifier":"Re-run the same stacking and forward modeling while replacing the stellar PSF with a PSF constructed from a different source class or measured only over the inner arcseconds, and vary the background subtraction by its uncertainty; if the net 50 kpc equivalent width changes sign or the halo scale lengths fall below about 5 kpc under any of those choices, the universality claim would not survive.","tokens_in":34025,"feed_emoji":"🌌","tokens_out":14226,"duration_ms":105756,"temperature":0.7,"pith_summary":"This paper stacks narrow-band images of 734 Lyman-$\\alpha$-selected galaxies and 119 continuum-selected galaxies at $z\\approx2$–3 to determine whether extended Lyman-$\\alpha$ halos are a generic property of star-forming galaxies. It argues that every subsample, including galaxies whose centers show net Lyman-$\\alpha$ absorption in slit spectroscopy, is a net Lyman-$\\alpha$ emitter on circumgalactic scales of $R \\gtrsim 50$ kpc, and that the spatial scale of the central Lyman-$\\alpha$ component matches the scale of the stellar continuum. If true, the halo-making process is common to all star-forming galaxies and their surrounding gas, with central absorption a small-scale feature embedded in a larger emitting envelope. The paper also shows that one two-exponential model convolved with an empirical point-spread function can describe both centrally emitting and centrally absorbing stacks.","feed_headline":"Stacks of 853 galaxies: all emit Lyman-alpha at 50 kpc","feed_subtitle":"Net absorbers in the center become net emitters in a 50 kpc aperture, making halos a generic galaxy feature.","key_machinery":"The load-bearing instrument is an empirical large-scale point-spread function built from 17–46 stars per field, represented by an analytic core-wing profile fit to each star and extrapolated beyond roughly 5 arcseconds, then smoothed and matched across fields so that a single effective PSF can be convolved with analytic surface-brightness models. Each Lyman-$\\alpha$ stack is fit with a seven-parameter forward model consisting of two exponentials (a central 'galaxy' component and an outer 'halo' component), a constant background, and center coordinates; the central amplitude is allowed to be negative, which is what lets the same model absorb or emit at the center. Parameterizing each exponential by its flux within $R=50$ kpc stabilizes the fit, and the derived fraction $f_{\\mathrm{gal}}^{\\mathrm{Ly}\\alpha}$ measures how much of the total Lyman-$\\alpha$ flux belongs to the central component. A parallel direct exponential fit to the one-dimensional profile reproduces the earlier literature method, and encircled equivalent-width curves turn the fitted profiles into an aperture-dependent statement about who counts as a Lyman-$\\alpha$ emitter.","core_discovery":"Using stacks of 734 Lyman-$\\alpha$-selected and 119 UV-continuum-selected galaxies in nine fields at mean redshift $z\\approx2.6$, the paper claims that extended Lyman-$\\alpha$ emission is universal: every stack, including the 55-galaxy stack with net central Lyman-$\\alpha$ absorption, has net positive Lyman-$\\alpha$ flux within a 50 kpc aperture. Direct exponential fits give halo scale lengths $r_0 \\approx 10$–21 kpc, several times the continuum scale lengths, and the forward-modeled halo scale lengths are $r_{0,\\mathrm{halo}} \\approx 9$–21 kpc. The fitted central Lyman-$\\alpha$ component has a scale length of roughly 1–3 kpc, consistent with the fitted continuum scale length, even in absorbing stacks, indicating that the absorbing gas is co-located with the UV-bright stars. The paper further finds only weak dependence of halo scale on galaxy luminosity or Lyman-$\\alpha$ equivalent width, with a slight tendency for continuum-faint, high-equivalent-width galaxies to have larger halos, in contrast to some earlier stacked surveys.","pith_inferences":["Editorial inference: The universality claim rests on the large-radius behavior of the stellar PSF; a PSF with slightly more power in its wings than the true extended-source profile would inflate halo scale lengths, so reconstructing the PSF from extended sources would be a sharper test.","Editorial inference: Because the analysis fixes one stacking recipe, the quoted uncertainties are formal lower bounds; varying the sigma-clipping threshold and flux scaling could move the scale lengths by more than the quoted 1-sigma errors.","Editorial inference: The near-constancy of halo scale across galaxy properties hints that environment, such as Mpc-scale overdensity, may matter more than luminosity or equivalent width; the paper's quasar-centered fields are plausibly overdense, a possibility it flags for future work.","Editorial inference: Literature differences between integral-field halo scales near 5 kpc and the 10–20 kpc scales found here and in other stacked surveys may stem from instrument- and method-dependent systematics, so cross-survey trends in halo size should not be read as purely physical until a common pipeline is applied to all data."],"forward_implications":["For every stacked subsample, Lyman-alpha light extends well beyond the UV continuum, with halo exponential scale lengths around 10–20 kpc regardless of central galaxy luminosity or equivalent width.","Galaxies classified as net Lyman-alpha absorbers in small apertures become net emitters within a 50 kpc aperture, so the emitter/absorber label is aperture-dependent rather than intrinsic.","The central Lyman-alpha component's fitted scale length matches the continuum scale length of roughly 1–3 kpc, even for absorbing stacks, placing the absorbing gas on the same spatial scale as the stars.","Encircled equivalent width grows by factors of 2–3 from the center out to 50–70 kpc, so slit-based classifications can underestimate the total Lyman-alpha output of a galaxy.","Continuum-faint, high-equivalent-width galaxies tend to have slightly larger halos, opposite to the trends reported by some earlier stacked surveys."],"supporting_citations":[{"why":"Supplies the earlier direct-exponential halo measurements and baseline scale lengths that this paper reanalyzes with point-spread-function corrections.","marker":"C. C. Steidel et al. (2011)"},{"why":"Provides the dual-exponential, PSF-convolved forward-modeling framework this paper adapts to stacks with central Lyman-alpha absorption.","marker":"F. Leclercq et al. (2017)"},{"why":"Supplies individual halo measurements and the two-component profile approach on which the forward model is based.","marker":"L. Wisotzki et al. (2016)"},{"why":"Defines the Lyman-alpha-selected sample, its selection, and its equivalent-width measurements used here.","marker":"R. F. Trainor et al. (2015, 2016)"},{"why":"Provides integral-field Lyman-alpha halo profiles of galaxies in the same survey fields, used for radial comparison.","marker":"Y. Chen et al. (2021)"},{"why":"Provides integral-field profiles of low-mass, high-ionization galaxies in the same fields, used as a comparison sample.","marker":"D. K. Erb et al. (2023)"},{"why":"Reports stacked halo trends with UV luminosity and equivalent width that this paper's weak trends contradict.","marker":"R. Momose et al. (2016)"},{"why":"Offers deep stacked double-exponential halo fits in one of the same survey fields, a direct comparison sample.","marker":"S. Kikuta et al. (2023)"}],"fun_headline_variants":["Even Lyman-alpha absorbers glow in 50 kpc halos","Universal Lyman-alpha halos: absorbers become emitters at 50 kpc","All galaxy stacks emit Lyman-alpha at 50 kpc, even absorbers","Lyman-alpha halos extend to 50 kpc for all galaxy types","Even net absorbers show Lyman-alpha emission in 50 kpc halos"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that even central Lyman-alpha absorbers are net emitters at 50 kpc assumes that the point-spread function built from stars, after smoothing and field matching, faithfully represents how extended Lyman-alpha light is spread, and that the sky background is flat and correctly subtracted at those radii.","fun_headline_variants_meta":{"raw":{"variants":["Even Lyman-alpha absorbers glow in 50 kpc halos","Universal Lyman-alpha halos: absorbers become emitters at 50 kpc","All galaxy stacks emit Lyman-alpha at 50 kpc, even absorbers","Lyman-alpha halos extend to 50 kpc for all galaxy types","Even net absorbers show Lyman-alpha emission in 50 kpc halos"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000534,"raw_usage":{"total_tokens":2646,"prompt_tokens":1102,"completion_tokens":1544,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":718,"completion_tokens_details":{"reasoning_tokens":1440}},"tokens_in":718,"tokens_out":1544,"duration_ms":10863,"temperature":1.0,"reasoning_tokens":1440,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T15:10:40.718444+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the same stacking and forward modeling while replacing the stellar PSF with a PSF constructed from a different source class or measured only over the inner arcseconds, and vary the background subtraction by its uncertainty; if the net 50 kpc equivalent width changes sign or the halo scale lengths fall below about 5 kpc under any of those choices, the universality claim would not survive.","supporting_citations":[],"review_version":1}