{"id":"c31634af-3d15-4be1-b8fc-d6f91ab47f1d","arxiv_id":"2505.17377","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Within the quasar proximity region at z~2.2, Ly-alpha emitters are about 5 sigma less dense than continuum-selected galaxies, implying quasar photoevaporation suppresses low-mass galaxy formation.","lead":"Quasars at z~2.2 are surrounded by fewer faint, young galaxies (Ly-alpha emitters) than by brighter, more massive galaxies, suggesting quasar radiation evaporates gas in small galaxies. This statistical signal from 18 stacked quasar fields supports the idea that quasars suppress galaxy formation in their immediate neighborhoods.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Photoevaporation claim hinges on the Figure 2 control; if the Harikane et al. halo-mass matching does not reproduce quasar environments, the 5-sigma LAE deficit could be environmental rather than caused by quasar feedback.","rationale":"The reader's weakest_assumption is exactly where I find the argument most exposed. The 5-sigma LAE-versus-continuum difference in Figure 1 is a well-constructed observed correlation, but the interpretation as photoevaporation requires the Figure 2 control to show that LAEs are suppressed around quasars relative to inactive galaxies of comparable halo mass. That demonstration inherits the systematics of the Harikane et al. (2022) mass relation and the assumption that quasar hosts and inactive hosts have identical environments. This is not an internal inconsistency in the stacking or significance estimates; it is an untested external assumption at the interpretation level. The abstract sentence about both populations having smaller densities near quasars is ambiguous and the EW>=150 A 'predominantly scarce' phrasing is overstated at ~2 sigma, but these are secondary. The conditional verdict remains appropriate pending the proposed mock-based test of the control assumption.","tokens_in":10404,"tokens_out":8994,"duration_ms":82893,"concrete_test":"Use a cosmological N-body simulation with an empirical galaxy and quasar prescription (e.g., UniverseMachine or similar, calibrated at z~2.2) to construct mock quasar fields. Place no-feedback quasars in halos matching the SDSS sample, apply the actual HSC/CLAUDS selection functions (NB387 LAE and photo-z continuum), and run the identical Figure 2 stacking: compare profiles around mock quasars with profiles around control galaxies selected by the same Harikane-relation halo-mass matching. If the no-feedback mocks reproduce the observed LAE deficit around quasars relative to control galaxies, the environmental explanation remains viable and Figure 2 does not exclude it. If the mocks predict flat, equal profiles, the environmental concern is mitigated and the feedback interpretation is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that the LAE deficit near quasars is caused by quasar radiation rather than by pre-existing environment. The only evidence separating these is the Figure 2 control, in which the same profiles are stacked around continuum-selected galaxies matched to the mean quasar halo mass via the Harikane et al. (2022) UV-luminosity-to-halo-mass relation. This control is load-bearing because, absent feedback, halos of the same mass should produce similar clustering. Two assumptions must hold: (1) the Harikane relation has negligible enough scatter that the selected galaxies genuinely occupy quasar-mass halos; (2) active quasars and inactive galaxies of the same halo mass trace identical large-scale environments. Neither is established. Abundance-matching relations carry ~0.2-0.3 dex scatter, and quasars are a transient, possibly merger-triggered subpopulation. If quasar hosts preferentially sit in lower-density regions or have different satellite populations (assembly bias), the Figure 2 difference would be expected without any feedback. The paper does not quantify how such selection effects alter the expected profiles. A related technical issue is that the many control centers are spatially correlated, so bootstrap errors on the control profile may be underestimated. Because the photoevaporation interpretation rests on this comparison, the raw 5-sigma LAE-versus-continuum profile alone cannot distinguish feedback from environment.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper stacks 18 SDSS quasars at z~2.2 in the HSC-SSP Deep/UltraDeep fields, using the proximity radius to rescale each field before stacking, and compares the radial surface density profiles of Ly-alpha emitters (LAEs) from Kikuta et al. (2023) with continuum-selected galaxies from Desprez et al. (2023). The main results are: (1) within the quasar proximity region (~5 arcmin), the LAE density is more than 5 sigma lower than that of continuum-selected galaxies; (2) high-EW LAEs (EW0 >~75 Angstrom) are less abundant than low-EW LAEs at ~3 sigma, and LAEs with EW0 >~150 Angstrom are especially scarce; and (3) both LAEs and continuum-selected galaxies are less clustered around quasars than around control galaxies selected to have similar halo masses. The authors interpret these findings as evidence for quasar photoevaporation of low-mass haloes, with weaker effects on more massive haloes.","tokens_in":10683,"tokens_out":3071,"duration_ms":28675,"significance":"If the interpretation is correct, this is one of the cleanest statistical demonstrations of quasar negative feedback at z~2.2, using a homogeneous, large-area dataset and a simple empirical stacking methodology. The paper benefits from using public catalogs and clearly describing the selection cuts. The central 5-sigma deficit between LAEs and continuum-selected galaxies is a robust empirical pattern, and the EW dependence is an interesting and potentially discriminating observable. However, the causal attribution to photoevaporation depends on the control experiment in Figure 2, which is not yet established at the level needed to exclude environmental differences.","major_comments":[{"comment":"The claim that quasar activity, rather than pre-existing environment, causes the LAE deficit hinges entirely on the control experiment in Figure 2. The control galaxies are selected to match the mean quasar halo mass via the Harikane et al. (2022) UV luminosity-to-halo-mass relation, but this relation has intrinsic scatter (~0.2-0.3 dex) and the paper does not test how that scatter, or possible assembly bias, affects the expected clustering amplitude around the control centers. If the control galaxies preferentially reside in different large-scale environments than quasar hosts, the difference in Figure 2 would arise without any feedback. Please quantify the sensitivity of the control comparison to plausible scatter in the M_h-M_UV relation and to the known transient nature of quasar host environments, for example by varying the scatter, by using a clustering-based halo mass estimate for the control sample, or by comparing the quasar autocorrelation function with that of the control galaxies.","section":"Section 4, Figure 2"},{"comment":"The stacking procedure normalizes each quasar field by the individual proximity radius, r_prox, so that all fields contribute with the same physical scale. This implicitly assumes that the radial distribution of the putative photoevaporation effect scales self-similarly with r_prox. If the efficiency of feedback does not scale linearly with the ionizing luminosity, or if quasar lifetimes vary with luminosity, the normalization could create or distort the observed profile shape. A simple test would be to split the sample by quasar luminosity (or by r_prox) and check that the stacked profile shape is invariant under the scaling; without such a test, the 5-sigma deficit within the normalized proximity region is not yet uniquely tied to the physical proximity effect.","section":"Section 3, Eq. (2) and normalization"},{"comment":"The paper acknowledges that the LAE and continuum-selected galaxy samples span different redshift path lengths (Delta z ~0.05 versus ~0.15) and that photo-z contamination affects the fainter continuum-selected galaxies, but it does not quantify the impact of these effects on the central comparison in Figure 1. Because the 5-sigma deficit is measured between two samples with different redshift selection functions, a null-feedback model could in principle produce a similar apparent deficit if, for example, the continuum-selected sample includes a radially varying contaminating population. Please provide a quantitative estimate (e.g., using the photo-z error distribution and the known redshift selection) or a simple mock catalog demonstrating that the measured difference cannot be explained by these selection effects alone.","section":"Section 4, last paragraph"},{"comment":"The bootstrap error bars for the control profiles in Figure 2 may be underestimated because the control galaxies are spatially correlated and multiple control centers fall within the same large-scale structure. The paper does not specify whether bootstrap resampling accounts for this correlation. I recommend using a block bootstrap over independent sightlines, or a jackknife over the 18 quasar fields, to verify that the apparent difference in clustering between quasars and control galaxies remains significant.","section":"Figure 2 and bootstrap errors"}],"minor_comments":[{"comment":"Typographical errors: 'calcurated' should be 'calculated' in the captions of Figures 1, 2, and 3, and 'devide' should be 'divide' in Section 3.","section":"Figure captions"},{"comment":"The sentence 'Figure 4 shows that the density of continuum-selected galaxies.' is incomplete and should be rephrased to state what the figure demonstrates.","section":"Section 4, Figure 4"},{"comment":"The selection 'cmodel i < 25.0' is not precisely defined; please state whether this is the i-band cmodel magnitude and clarify the completeness limit in the context of the Desprez et al. (2023) catalog.","section":"Section 2.2"},{"comment":"The axis labels mix arcmin and pMpc without clearly explaining that the physical scale is computed for the median proximity radius; please add a note clarifying that the pMpc scale is approximate and relies on the assumed cosmology.","section":"Figure 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a concise and potentially important result, but the central physical conclusion hinges on the Figure 2 control, which is not yet convincing. The authors should be encouraged to strengthen the control analysis and to explicitly propagate the scatter in the halo-mass matching and the redshift-selection differences. If the control concern is resolved, the paper would be a strong contribution to the study of quasar feedback."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Take a look at Suzuki et al. if you want the current state of the quasar-photoevaporation question. The new thing here is that they stack 18 SDSS quasars at z~2.2 and compare two galaxy populations in the same fields: LAEs (low-mass, high-EW) and continuum-selected galaxies (higher-mass). The headline result is that LAEs are suppressed by >5 sigma relative to continuum galaxies inside the quasar proximity region, and that high-EW LAEs are scarcer than low-EW ones. That differential measurement is genuinely new and is built from external catalogs with a straightforward, reproducible stacking procedure. I think that part holds up.\n\nThe paper also adds a control: they stack the same galaxy samples around continuum-selected galaxies whose halo masses are matched to the quasars via the Harikane et al. (2022) luminosity-halo-mass relation, and find weaker clustering around the quasars. This is where the interpretation moves from 'LAEs avoid quasar vicinity' to 'quasar activity suppresses galaxy formation.' The control is clever, but it is load-bearing, and it inherits every assumption in that mass relation. Scatter of 0.2-0.3 dex in abundance matching, plus the possibility that active quasar hosts are not a random draw from halos of that mass, could produce exactly this kind of environmental difference without any feedback. The paper does not quantify how much scatter would blur the control, and the many control centers are spatially correlated, so the bootstrap errors likely underestimate the uncertainty. The stress-test note is right about this.\n\nTwo smaller issues: the EW>=150 A result is about 2 sigma, not the 'predominantly scarce' the abstract implies; and the last sentence of the abstract about both populations having smaller densities closer to quasars is confusing, since Figure 1 shows continuum galaxies rising toward the quasar. It presumably refers to the control comparison and should be reworded.\n\nNone of this kills the paper. The central LAE-versus-continuum deficit is a clean observation, the EW trend is suggestive, and the authors are honest about the photo-z and path-length limitations. If I were the editor I would send it out, but I would ask the referees to focus on the control: can the authors show how robust the Figure 2 result is to scatter in the mass relation, and can they test the environmental hypothesis directly? This is a useful contribution to a real controversy, not a settled claim.","headline":"A solid new differential measurement of LAE deficits around quasars, but the feedback interpretation leans on a halo-mass-matched control whose assumptions are not yet nailed down.","tokens_in":11207,"tokens_out":4016,"would_cite":true,"duration_ms":44449,"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":"Quasars at z~2.2 leave a >5-sigma deficit of young galaxies within ~2.5 pMpc, pointing to photoevaporation.","keywords":["quasars","photoevaporation","Ly-alpha emitters","galaxy environments","negative feedback","z~2.2","HSC-SSP","halo mass"],"falsifier":"Stack a comparable sample of quasars at $z\\sim2.2$ with deep spectroscopy of LAE candidates inside the proximity region: if LAEs with rest-frame EW above 150 Å are found within about 1 pMpc of active quasars, or if the density deficit disappears once photometric-redshift interlopers are removed, the photoevaporation claim would be contradicted.","tokens_in":10236,"feed_emoji":"🔭","tokens_out":5909,"duration_ms":39922,"temperature":0.7,"pith_summary":"This paper attempts to show that ultraviolet radiation from a quasar can photoevaporate gas in nearby low-mass dark matter halos, suppressing or hiding galaxy formation in the immediate vicinity. By stacking 18 quasar fields at $z\\sim2.2$ and comparing the radial density of Ly$\\alpha$ emitters (LAEs, small young galaxies) with continuum-selected galaxies (larger, more massive ones), the authors find that LAE density is more than $5\\sigma$ lower than continuum-selected galaxy density within the quasar proximity region of roughly 2.5 pMpc. High equivalent-width LAEs are disproportionately scarce, and both populations become less dense closer to the quasar. If correct, this is statistical evidence that quasars exert negative feedback on surrounding galaxies, helping to explain why quasars are not always found in galaxy overdensities.","feed_headline":"Quasar radiation suppresses nearby small galaxies at >5 sigma","feed_subtitle":"Stacking 18 quasar fields at z~2.2 shows Ly-alpha emitters vanish inside the proximity zone while larger galaxies remain.","key_machinery":"The analysis rests on stacking quasar fields with spatial scales normalized so that each quasar's proximity radius, computed from its Lyman-limit luminosity and the assumed UV background intensity ($J_{21}=1.0$), maps to the median value of 5.2 arcmin (about 2.5 pMpc). Within the stacked fields, the surface densities of LAEs and continuum-selected galaxies are measured in radial bins and normalized at 20 arcmin. The differential between the two populations, the split of LAEs into high- and low-EW subsamples at EW$_0=75$ Å, and a control experiment using continuum-selected galaxies with quasar-matched halo masses are what carry the photoevaporation argument: the LAE deficit is strongest precisely where quasar radiation dominates over the UV background.","core_discovery":"The central claim is that quasar UV radiation suppresses the formation or visibility of low-mass galaxies within a few pMpc of the quasar at $z\\sim2.2$, with the effect strongest for the smallest halos. The evidence is a stacked analysis of 18 SDSS quasars in HSC-SSP Deep/UltraDeep fields: after normalizing each field by its proximity radius, LAEs are $\\gtrsim5\\sigma$ less dense than continuum-selected galaxies inside the proximity region, high-EW LAEs are about $3\\sigma$ less dense than low-EW ones, and LAEs with rest-frame EW $\\gtrsim150$ Å are almost entirely absent. The authors also show that both LAEs and continuum-selected galaxies are more clustered around halo-mass-matched control galaxies than around quasars, which they interpret as a signature of quasar activity rather than a pre-existing environmental difference.","pith_inferences":["A direct test of the photoevaporation interpretation would compare LAE deficits around quasars of different UV luminosities: if the deficit scales with ionizing flux or quasar lifetime, feedback is the likely cause; if not, an evolved-galaxy explanation becomes more plausible.","The same stacked approach could be applied at $z>6$, where the UV background is weaker and halos are less massive, predicting an even stronger LAE deficit around the earliest quasars.","Spectroscopic follow-up of continuum-selected galaxies inside the proximity region could separate photoevaporation from environmental effects: if their stellar masses are similar to those outside, the deficit is less likely to be a mass-dependent selection artifact."],"forward_implications":["Quasar activity at $z\\sim2.2$ should leave a measurable deficit of low-mass galaxies within roughly 2.5 pMpc, not just of LAEs.","High-EW LAEs act as a diagnostic of active quasar feedback: their scarcity inside the proximity region marks where UV radiation has stripped or heated halo gas.","The inferred halo-mass threshold for photoevaporation lies near $M_h\\sim3\\times10^9\\,M_\\odot$, consistent with theoretical delay-time estimates for gas removal.","The effect may extend to halos as massive as about $10^{12}\\,M_\\odot$, implying quasar feedback shapes galaxy populations over a wide mass range.","Environment studies that rely only on LAEs will systematically undercount neighbors around quasars, biasing clustering measurements if the effect is real."],"supporting_citations":[{"why":"Supplies the large SILVERRUSH LAE catalog at z=2.2 that the LAE density measurements are built on.","marker":"Kikuta et al. (2023)"},{"why":"Supplies the continuum-selected galaxy catalog with photometric redshifts used as the comparison population.","marker":"Desprez et al. (2023)"},{"why":"Provides the average halo mass of SDSS quasars used to match control galaxies to quasar environments.","marker":"Eftekharzadeh et al. (2015)"},{"why":"Provides the UV luminosity-to-halo-mass relation used to assign halo masses to continuum-selected galaxies for the control experiment.","marker":"Harikane et al. (2022)"},{"why":"Provides the photoevaporation delay-time model and the method for estimating the quasar proximity region size.","marker":"Kashikawa et al. (2007)"},{"why":"Provides the hydrodynamical simulations of photoevaporation that set the halo-mass and UV-intensity thresholds for gas stripping.","marker":"Kitayama et al. (2000, 2001)"},{"why":"Gives the measured UV background intensity at the Lyman limit used to normalize the proximity radii.","marker":"Cooke et al. (1997)"},{"why":"Reports the earlier finding of scarce high-EW LAEs near quasars that this work extends with simultaneous continuum-selected galaxies.","marker":"Uchiyama et al. (2019)"},{"why":"Shows that intense UV radiation can suppress star formation in massive halos, supporting the interpretation of reduced continuum-selected galaxy density.","marker":"Bruns et al. (2012)"}],"fun_headline_variants":["Quasar UV light suppresses small galaxies at z~2.2","Stacking 18 quasars shows Ly-alpha emitters vanish nearby","Small galaxy density drops >5 sigma inside quasar proximity","Photoevaporation by quasars erases tiny galaxy halos","Quasar feedback quenches low-mass galaxy formation nearby"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that quasar activity causes the observed deficit assumes that the continuum-selected galaxies used as controls really have the same halo masses and live in the same large-scale environments as the quasars; if the halo-mass matching fails, the weaker clustering around quasars could reflect a pre-existing environmental difference rather than feedback.","fun_headline_variants_meta":{"raw":{"variants":["Quasar UV light suppresses small galaxies at z~2.2","Stacking 18 quasars shows Ly-alpha emitters vanish nearby","Small galaxy density drops >5 sigma inside quasar proximity","Photoevaporation by quasars erases tiny galaxy halos","Quasar feedback quenches low-mass galaxy formation nearby"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000445,"raw_usage":{"total_tokens":2284,"prompt_tokens":1016,"completion_tokens":1268,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":632,"completion_tokens_details":{"reasoning_tokens":1181}},"tokens_in":632,"tokens_out":1268,"duration_ms":13929,"temperature":1.0,"reasoning_tokens":1181,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:47:39.992182+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Stack a comparable sample of quasars at $z\\sim2.2$ with deep spectroscopy of LAE candidates inside the proximity region: if LAEs with rest-frame EW above 150 Å are found within about 1 pMpc of active quasars, or if the density deficit disappears once photometric-redshift interlopers are removed, the photoevaporation claim would be contradicted.","supporting_citations":[{"cited_title":"J., Espey, B., & Carswell, R","cited_arxiv_id":null,"evidence_quote":"Gives the measured UV background intensity at the Lyman limit used to normalize the proximity radii."}],"review_version":1}