{"id":"99358184-2e3b-4a36-be61-1cdc91469be5","arxiv_id":"2507.13456","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A JWST medium-band survey finds 119 Hβ+[OIII] emitters at z~7, including 19 with rest-frame equivalent widths over 3000 Å, and links spatially offset line emission to massive-star-driven outflows.","lead":"Using JWST medium-band images, the authors identify 119 galaxies at redshift about 7 whose light is dominated by hydrogen and oxygen emission lines, 19 of them with extreme equivalent widths beyond 3000 angstroms. The paper uses the spatial position of the line emission to argue that many of these extreme emitters are powered by young massive stars with strong outflows, not by black holes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Offset claim rests on an untested line-map bias at the 0.06 arcsec scale; a simulation of no-offset galaxies with realistic color gradients and astrometric residuals should be run before the shock-heating interpretation is accepted.","rationale":"After reading the paper in good faith, I find the sample selection, photometric SED fitting, and the existence of a population with extreme EW0 are credible: the 28-object spec-z validation in Appendix A, the clear color-excess selection, and the consistency with external high-EW samples all support those results. The load-bearing claim is not the existence of the emitters but the physical inference about shock-heated outflows from 13 spatially offset line distributions. That inference collapses if the offsets are systematic artifacts. The reader's verdict (CONDITIONAL) is appropriate: the central claim should be accepted only after the line-map bias is tested. My concern refines rather than replaces the reader's: the intrinsic-color term is probably small in f-nu, but the 0.06 arcsec scale makes PSF-matching residuals, astrometric alignment, and spatially varying continuum color the real risks, and none are quantified. An end-to-end simulation, or a re-measurement using per-pixel SED-based continuum subtraction, would settle whether the 13 detections survive. Therefore the verdict should remain CONDITIONAL with the requested check.","tokens_in":26773,"tokens_out":13276,"duration_ms":153066,"concrete_test":"Run the paper's exact pipeline (PSF-matching, Eq. 7 line-map construction, DAOStarFinder centroiding, and the d > 0.3 kpc criterion) on simulated z ~ 7 EELGs with no intrinsic line offset: adopt the median SED (beta = -2.2, Av = 0.1) and a Sersic profile, add a smooth spatial F410M minus F444W continuum color gradient (for example +/- 5 percent across the half-light radius) and/or WCS misalignments up to 0.03 arcsec between F150W/F356W and F444W, and inject realistic noise. Measure the false-positive fraction of d(Line - UV) > 0.3 kpc over about 1000 realizations. If the false-positive rate exceeds roughly 5 percent, the 13/119 offsets cannot support the shock-heating claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Sec 5's central claim—13 H beta plus [OIII] emitters with spatially offset line emission 'unequivocally' caused by massive-star-driven shocks—depends wholly on the fidelity of the Eq. 7 line map at the 0.3 kpc scale, which corresponds to roughly 0.06 arcsec at z ~ 7, near the NIRCam PSF FWHM and typical astrometric/PSF-matching residuals. Sec 3.4 explicitly leaves the intrinsic F410M - F444W continuum color out of the map and asserts without a quantitative test that it does not prominently affect the result. In f-nu units the median beta ~ -2.2 continuum is nearly flat (about 2 percent difference) between the filters, so the flat-color assumption is less dangerous than a Balmer or 4000 Angstrom break, which is not at rest 0.5 micron; however, any spatially varying color gradient, residual PSF mismatch after convolution, or inter-filter WCS shift (for example F150W/F356W versus F444W) at the tens-of-milliarcsecond level would shift line centroids by more than 0.3 kpc and create the 13 offset detections. The quoted centroid errors (0.02 to 0.04 kpc in Table 5) come from varying initial guesses only, not from photon noise, PSF uncertainty, or astrometric registration, so they do not bound the systematic error. The paper's own Sec 5.3 further concedes that offsets do not rule out AGN/HII photoionization, softening the unequivocal language of Sec 5.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses JWST Cycle-1 medium-band F410M and broad-band F444W imaging to identify 119 Hβ+[OIII] emitters at z~7 in the CEERS, PRIMER, PAR1199, NEP, and JADES fields. The authors measure rest-frame equivalent widths (420 < EW0 < 6850 Å), stellar masses, UV slopes, dust attenuation, sizes, and clumpy fractions, and single out 19 high-EW (EW0>3000 Å) sources. They present spatially resolved line maps from the F410M-F444W excess and report 13 objects whose Hβ+[OIII] emission is offset by >0.3 kpc from the UV and stellar continuum. On this basis, together with a model-dependent upper limit on AGN contribution, they conclude that these emitters are likely powered by radiative shocks from massive-star-driven outflows. They also report a strong overdensity of EELGs in JADES GOODS-S at z~7.3.","tokens_in":27007,"tokens_out":9848,"duration_ms":99747,"significance":"If the spatial-offset interpretation is correct, the paper provides a novel observational handle on the ionizing source in extreme emission-line galaxies at the epoch of reionization, suggesting that some z~7 EELGs are starbursts with feedback-driven outflows rather than AGN-dominated systems. The sample construction is careful in several respects: the F090W-dropout plus F410M excess selection is explicit, and Appendix A validates the photometric redshifts against 28 spectroscopic redshifts. The 19 high-EW objects are rare and well characterized in mass, color, and dust content. However, the headline shock-heating claim currently rests on centroid offsets measured from a line map whose zero point is not tested against continuum color gradients, PSF-matching residuals, or astrometric registration errors. The paper's own Section 5.3 concedes that the offsets do not rule out photoionization, which is in tension with the 'unequivocal' language used in Section 5 and the Abstract. With a proper null test and appropriately softened claims, this would be a valuable contribution; in its present form, the central physical interpretation is not yet fully supported.","major_comments":[{"comment":"The emission-line map Fline is constructed by subtracting F444W from F410M after PSF matching, assuming the continuum is flat between the two filters. The text states that the intrinsic color is not taken into account and that it does not 'prominently affect' the result, but no quantitative test is provided. At z~7, the 0.3 kpc offset threshold corresponds to roughly 0.06 arcsec, which is comparable to the NIRCam PSF core and to typical inter-filter astrometric residuals. The centroid uncertainties quoted in Table 5 and Section 4.3 are derived by varying initial guesses only; they do not include photon noise, PSF mismatch, or WCS registration errors. A null test with simulated no-offset galaxies that include realistic continuum color gradients, PSF-matching residuals, and astrometric shifts is required before the 13 offset detections can be used as evidence for shock heating.","section":"Sec. 3.4, Eq. (7)"},{"comment":"The claim that the observed offsets 'can be attributed unequivocally to the third scenario' (shock heating) is internally contradicted by Section 5.3, which states that the offset does not rule out photoionization from AGN or H ii regions. Because line-of-sight projection cannot be distinguished and ionized gas can be spatially offset from the continuum in star-forming regions with patchy dust or escaping radiation, the data support 'consistent with shock heating' rather than 'unequivocal.' The Abstract's statement that these emitters 'are likely under strong feedback-driven winds from massive stars' should be softened to a candidate interpretation unless the null test in the previous comment is provided. In addition, the 'absence of obvious signatures of actively accreting black holes' rests on the model-dependent upper-limit argument of Section 5.2, not on direct AGN diagnostics.","section":"Sec. 5 and Abstract"},{"comment":"As printed, the selection inequality mF410M−mF444W > −2.5 log10(1−3 sqrt(σ^2_F410M+σ^2_F444W)/fF410M) has a positive right-hand side for typical σ/f (e.g., ~0.39 for σ/f=0.1), so it selects sources with fF410M < fF444W, i.e., objects red in F410M−F444W. This is the opposite of the intended line-excess selection and is inconsistent with Eq. (3) and Figure 2, where the emitters lie at negative F410M−F444W. The equation or its accompanying explanation should be corrected so that the published selection criterion reproduces the sample; if the actual code used the correct inequality, that should be stated explicitly.","section":"Sec. 3.1, Eq. (2)"}],"minor_comments":[{"comment":"The offset criterion is printed twice as 'd(Line − UV) > 0.3 kpc and d(Line − UV) > 0.3 kpc'; the second condition should presumably be d(Line − Stellar) > 0.3 kpc. Please correct the typo in both places.","section":"Sec. 4.3 and Sec. 5.3"},{"comment":"The F410M−F444W threshold in Eq. (3) corresponds to rest-frame EW of roughly 660 Å, while the paper defines EELGs as EW0 > 1000 Å. Please clarify how the selection boundary maps onto the final EELG definition and why some objects with EW0 below 660 Å appear in the final sample.","section":"Sec. 3.1"},{"comment":"The PSF matching between F410M and F444W is mentioned but the matching procedure, kernel size, and any validation of the residual PSF differences are not described. A brief description or reference would aid reproducibility and help assess the impact of PSF mismatch on the offset measurements.","section":"Sec. 3.4"},{"comment":"The reported overdensity δgal ~ 50 within a 240 kpc diameter aperture should be accompanied by an estimate of the expected Poisson variance. If the average number of EELGs per such aperture, ⟨ngal⟩, is of order unity, the quoted overdensity may be strongly affected by shot noise.","section":"Sec. 6"},{"comment":"There are several typographical and notation issues, including 'EEGLs' in the Abstract (should be 'EELGs'), inconsistent 'Hβ+[Oiii]' vs. 'Hβ+[OIII]' in figure captions, and the sentence in Section 6 that 'the average value is less than the standard deviation in all regions,' which is unclear and should be reworded.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a valuable sample and a potentially interesting spatial-offset signal, and the core dataset appears to be carefully constructed. However, the central physical interpretation is currently pushed beyond what the imaging analysis can support without a null test of the line-map systematics. The sign error in Eq. (2) also needs to be fixed before the sample selection can be reproduced. If the authors add a realistic simulation of no-offset galaxies, correct the selection equation, and temper the 'unequivocal' language, the paper could be acceptable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. First, the sample is real and useful: 119 Hβ+[OIII] emitters at z~7 selected from public JWST medium-band imaging, with a well-documented two-step selection (Lyman-break plus F410M excess), a 28-object spec-z sanity check in Appendix A, and 19 objects with rest-frame EW > 3000 Å. The EW distribution, low masses, blue UV slopes, and low dust content are all derived with standard SED fitting and look solid. This is the first statistical sample of these extreme emitters at z~7, and the line-size and centroid measurements, even if some are later revised, are a step forward. The clumpy fraction comparison and the overdensity in JADES-GOODS-S are also worth having; the overdensity factor of ~50 is quoted without an error bar, but the same region was spectroscopically confirmed by Helton et al. 2023, so the existence of an overdensity is not in doubt.\n\nSecond, the headline interpretation—that 13 offset emitters are 'unequivocally' powered by radiative shocks from massive-star outflows—does not yet hold up. The line maps in Eq. 7 subtract F444W from F410M without accounting for intrinsic continuum color, and the paper explicitly says so. At z~7, 0.3 kpc is about 0.06 arcsec, close to the NIRCam PSF core and typical astrometric/PSF-matching residuals. The flat-color assumption is less dangerous than it sounds because the median β~−2.2 continuum gives only a ~2% color difference between the two filters, but spatially varying color gradients, residual PSF mismatch, or inter-filter WCS shifts at the tens-of-milliarcsecond level could easily produce offsets of this size. The quoted centroid errors in Table 5 come from varying initial guesses only, not from photon noise, PSF uncertainty, or astrometric registration, so they do not bound the systematic error. The paper itself softens the claim in Sec 5.3, noting that offsets do not rule out AGN or HII photoionization, which sits awkwardly with the 'unequivocally' in Sec 5.\n\nMy take: the sample and the extreme-EW population are likely robust, and the AGN upper-limit argument (using external scalings, not a fit) is reasonable as a bound. The offset/shock claim needs a null test—simulate no-offset galaxies with realistic color gradients and astrometric residuals, run them through the same line-map construction, and see how many spurious offsets appear. That should be a required revision, not an optional extra. The paper deserves serious peer review, but with that test as a condition. I would certainly cite the sample and would bring it to a reading group for a lively discussion of systematics.","headline":"A genuinely useful z~7 EELG sample, but the shock-heating claim rests on an untested 0.06 arcsec centroid systematics; needs a no-offset simulation before it can carry weight.","tokens_in":27770,"tokens_out":1859,"would_cite":true,"duration_ms":24605,"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":"Thirteen of the most extreme emission-line galaxies at z≈7 show Hβ+[OIII] emission offset from their starlight, pointing to massive-star-driven outflows rather than AGN.","keywords":["extreme emission line galaxies","epoch of reionization","H beta + [OIII] emitters","JWST medium-band imaging","equivalent width","shock heating","galaxy outflows","galaxy overdensity"],"falsifier":"A simulation that injects realistic z≈7 galaxies with a Balmer/4000 Å break gradient into the F410M/F444W pair and runs the same PSF-matching and centroid pipeline could show whether a significant fraction of the 13 offset sources appear offset in the absence of any outflow; if so, the shock-heating claim would collapse. A cheaper check is integral-field spectroscopy of the 13 sources: if the gas kinematics show no broad or shifted component consistent with an outflow, the shock interpretation is unsupported.","tokens_in":26443,"feed_emoji":"🌌","tokens_out":4183,"duration_ms":43470,"temperature":0.7,"pith_summary":"The paper attempts to establish that the most extreme emission-line galaxies at redshift z≈7, identified by a flux excess in the JWST medium-band filter F410M, are powered mainly by young starbursts and massive-star-driven outflows rather than by active galactic nuclei. The case rests on 119 Hβ+[OIII] emitters, 19 of which have rest-frame equivalent widths above 3000 Å. For 13 of these, the line emission is spatially offset by more than 0.3 kpc from the rest-frame UV and stellar continuum, an arrangement the paper argues cannot arise from photoionization by stars or AGN and therefore points to radiative shocks in winds from massive stars. If correct, this population contributes to cosmic reionization through ionizing photons and large ionizing bubbles, not through black hole accretion.","feed_headline":"13 reionization-era galaxies show starlight-offset emission","feed_subtitle":"JWST medium-band maps point to massive-star outflows, not black holes, driving extreme line widths.","key_machinery":"The central tool is the medium-band flux excess technique: F410M contains Hβ+[OIII] at z≈7 while F444W samples the underlying continuum. The line map is built by $F_{\\rm line} = (f_{\\rm F410M} - f_{\\rm F444W})/(1 - \\Delta_{\\rm F410M}/\\Delta_{\\rm F444W}) \\times \\Delta_{\\rm F410M}$ with F410M PSF-matched to F444W, and the same pair of images yields a continuum map via Equation (8). The centroid-offset analysis then compares the line map with F150W (rest-frame UV) and F356W (stellar continuum) using peak finding and DAOStarFinder.","core_discovery":"Using PSF-matched F410M and F444W images, the authors build continuum-subtracted maps of Hβ+[OIII] and measure the centroid offsets between the line, the rest-frame UV (F150W), and the stellar continuum (F356W). They find 13 of 119 emitters at z≈7 where the line centroid is offset by more than 0.3 kpc from both UV and stellar light while UV and stellar centroids coincide. Because photoionization by either stars or AGN would keep these components aligned, the paper concludes that these offsets reveal shock-heated gas in outflows driven by massive-star feedback. The same argument leads to the claim that high-EW emitters (EW0 > 3000 Å), typically low-mass (log M★/M⊙ ≲ 7.5), blue (β ~ −2.2), and dust-poor (AV ~ 0.1), are starbursting systems that may be key sources of ionizing photons during reionization.","pith_inferences":["A direct prediction of the shock-heating interpretation is that the 13 offset sources should show broad or shifted gas kinematics in integral-field follow-up, with line ratios such as [OIII]/Hβ enhanced over pure photoionization predictions.","The method could be extended to lower equivalent-width emitters at z≈7 to test whether the offset fraction scales with EW; such a trend would link shock heating specifically to the most extreme starbursts.","If the z≈7.3 overdensity also hosts the nearby quiescent galaxy JADES-GS-z7-01-QU, feedback from the clustered starbursts may be quenching neighbors, a testable connection between reionization-era overdensities and early quiescence.","The 0.3 kpc offset threshold is comparable to the NIRCam PSF core at these wavelengths, so a version of this analysis with grism or higher-resolution data could verify whether some apparent offsets are PSF-matching artifacts."],"forward_implications":["If the line offsets are genuine, some of the most extreme EELGs at z≈7 are not AGN but starbursts whose winds ionize gas on kiloparsec scales, changing the inferred ionizing-photon budget for reionization.","At least 11% (13 of 119) of z≈7 EELGs would be powered substantially by shock heating, a contribution usually neglected in reionization models.","The observed low masses, blue UV slopes, and low dust attenuation show that EW0 > 3000 Å can be achieved without invoking a top-heavy IMF or AGN, supporting bursty star-formation histories as the dominant channel.","The EELG overdensity in JADES GOODS-S, coincident with a spectroscopically confirmed z = 7.265 structure, suggests that large ionizing bubbles form around clustered EELGs, aiding reionization in dense environments.","The centroid-offset method can be applied to other medium-band surveys to map stellar feedback in reionization-era galaxies without spectroscopy."],"supporting_citations":[{"why":"Supplies the public JWST datasets, the dropout selection, and the PSF-matching procedure on which the emitter catalog and line maps are built.","marker":"Morishita et al. (2023)"},{"why":"Provides the spectroscopic confirmation of the z = 7.265 overdensity with δgal = 6.631, anchoring the paper's claim of an EELG overdensity.","marker":"Helton et al. (2023)"},{"why":"Offers the precedent of a z = 7.28 high-EW Lyα emitter with high escape fraction, used to argue that high-EW emitters are strong reionization contributors.","marker":"Saxena et al. (2023)"},{"why":"Supplies the observed faint AGN fraction at 6 < z < 9, used to set the MBH/M⋆ assumption in the AGN contribution upper limit.","marker":"Harikane et al. (2023)"},{"why":"Reports a similar conclusion that high-EW emitters are unlikely to be dominated by AGN, supporting the paper's non-AGN interpretation.","marker":"Zhang et al. (2023)"}],"fun_headline_variants":["13 z≈7 emitters show offset gas: massive-star winds, not AGN","JWST: extreme emitters at z≈7 are starbursting, not AGN","Offset emission in 13 reionization galaxies points to winds","Starburst outflows, not black holes, explain extreme z≈7 galaxies","13 z≈7 galaxies: offset emission from starburst winds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The interpretation assumes the continuum-subtracted line map is not shifted by the unmodeled intrinsic color difference between F410M and F444W; the paper states this color is not taken into account and asserts, without a quantitative test, that it does not prominently affect the centroid offsets.","fun_headline_variants_meta":{"raw":{"variants":["13 z≈7 emitters show offset gas: massive-star winds, not AGN","JWST: extreme emitters at z≈7 are starbursting, not AGN","Offset emission in 13 reionization galaxies points to winds","Starburst outflows, not black holes, explain extreme z≈7 galaxies","13 z≈7 galaxies: offset emission from starburst winds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001303,"raw_usage":{"total_tokens":5418,"prompt_tokens":1152,"completion_tokens":4266,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":768,"completion_tokens_details":{"reasoning_tokens":4163}},"tokens_in":768,"tokens_out":4266,"duration_ms":38633,"temperature":1.0,"reasoning_tokens":4163,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:25:04.064137+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A simulation that injects realistic z≈7 galaxies with a Balmer/4000 Å break gradient into the F410M/F444W pair and runs the same PSF-matching and centroid pipeline could show whether a significant fraction of the 13 offset sources appear offset in the absence of any outflow; if so, the shock-heating claim would collapse. A cheaper check is integral-field spectroscopy of the 13 sources: if the gas kinematics show no broad or shifted component consistent with an outflow, the shock interpretation is unsupported.","supporting_citations":[],"review_version":1}