{"id":"fa4a0369-dd19-417e-ba84-b85f019fd18f","arxiv_id":"1908.02294","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A compact, massive galaxy at z=2.234 shows lower molecular gas fractions and shorter depletion times in its central kiloparsecs, the first resolved CO(1-0) evidence for inside-out quenching at high redshift.","lead":"Using high-resolution VLA images of carbon monoxide, researchers mapped the molecular gas inside a massive, compact galaxy at redshift 2.2. They found the gas is depleted fastest in the galaxy's center, consistent with star formation shutting off there first, a process called inside-out quenching.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central f_H2 suppression may be inflated by an unresolved AGN continuum in the F160W band; Section 3.2 considers only stellar M/L gradients and never tests for a central point source, leaving the inside-out quenching claim vulnerable.","rationale":"I read the paper in good faith: the VLA CO(1-0) imaging is a substantial observational achievement, the visibility-based fitting is appropriate for marginally resolved sources, and the authors are transparent about conversion-factor assumptions and the need for no gas replenishment. Those admitted limitations justify the reader's CONDITIONAL verdict. However, the weakest point is not only the global conversion factors; it is the untreated possibility of AGN light in the F160W profile. The galaxy belongs to a class with elevated AGN fractions, the X-ray limit is weak for obscured AGN, and Section 3.2's enumeration of systematic effects omits non-stellar contamination. If even about 20 percent of the central F160W flux is AGN continuum, the stellar mass profile becomes artificially concentrated and the central f_H2 suppression could be reduced or erased. The proposed point-source decomposition is a direct, feasible test with archival data. Because this concern is an additional source of uncertainty rather than a demonstrated error, it does not change the verdict: CONDITIONAL remains appropriate, but the revision should address AGN contamination explicitly.","tokens_in":17997,"tokens_out":10848,"duration_ms":132354,"concrete_test":"Run a two-component GALFIT decomposition of the HST WFC3/F160W image allowing a Sersic profile plus a PSF point source, and examine the residual map for a central unresolved excess at >3 sigma. If a point-source component is detected, subtract it from the stellar mass profile, recompute f_H2 with the same CO map, and test whether the central suppression (f_H2 rising by a factor of roughly 2 from the central 1-2 kpc to 2-4 kpc) persists. If the suppression disappears, the inside-out quenching evidence is not secure.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim rests on the radially resolved f_H2 profile, which combines CO(1-0) with an F160W-derived stellar mass map under a single mass-to-light ratio (Section 2.4). Section 3.2 discusses radial M/L variations but only in the direction that strengthens the central suppression (redder centers with higher M/L) and does not consider non-stellar AGN continuum. COSMOS 27289 is a compact SFG, a population with a high AGN incidence (Barro et al. 2013; Kocevski et al. 2017), and the current X-ray nondetection does not exclude a Compton-thick AGN. If a faint AGN contributes to the central F160W flux, the stellar mass profile is artificially more peaked, the central f_H2 is pushed down, and the observed inside-out gradient could be partly or wholly an artifact. The same AGN could heat dust and bias the 270 micron SFR profile, further affecting t_dep. The paper never performs a point-source decomposition or discusses this alternative, so the central observation is not yet isolated from AGN contamination.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents new VLA B-array CO(1-0) imaging of the z=2.234 compact star-forming galaxy COSMOS27289, combined with previously published C-array data and archival HST/WFC3 F160W and ALMA 343 GHz images. The authors extract elliptical Sersic radial profiles (GALFIT for HST, a visibility-based MCMC fitter for VLA and ALMA), deconvolve the PSF or synthesized beams, and apply a first-order residual correction following Szomoru et al. (2010). Under assumptions of a global F160W mass-to-light ratio, a single alpha_CO = 1, and a linear 270 micron-to-SFR conversion, they derive radial profiles of the molecular gas fraction f_H2 and depletion time t_dep. These profiles are lower in the central 1-2 kpc and rise at radii ~2-4 kpc, which the authors interpret as evidence for imminent inside-out quenching, the first such claim from resolved molecular gas observations at high redshift. They compare the observed f_H2 profile to the Tacchella et al. (2016) simulations and report good qualitative and semi-quantitative agreement.","tokens_in":18215,"tokens_out":9339,"duration_ms":100005,"significance":"If the central suppression is real, this is an important and rare direct measurement of the molecular gas distribution during the quenching of a compact high-redshift galaxy. The choice of ground-state CO(1-0) is well motivated: it is less affected by dust extinction and AGN than rest-UV or H-alpha tracers, and the difference with the higher-J CO observations of Barro et al. (2017) is clearly discussed. The technical approach is appropriate: visibility-based Sersic fitting is well suited to marginally resolved interferometric data, and the MCMC uncertainties are propagated through to the derived profiles. The central result is not circular: the radial trend is directly observed in ratioed imaging under stated conversion assumptions, and the simulation comparison is not used to calibrate the observations. The paper also systematically argues that likely radial variations in the stellar M/L, alpha_CO, and dust temperature would accentuate, not erase, the central suppression. The main caveats are the simple conversion factors, the marginal resolution of the ALMA data, and the lack of any test for a central AGN point-source contribution to the F160W light.","major_comments":[{"comment":"The F160W profile is used as the stellar mass map, but the paper does not test whether a central AGN contributes to the rest-frame ~500 nm light. COSMOS27289 is a compact SFG, a population with a high AGN incidence (Barro et al. 2013; Kocevski et al. 2017), and the X-ray nondetection does not exclude a Compton-thick AGN. If a faint unresolved source contributes to the central F160W flux, the stellar mass profile is artificially more centrally peaked, which would lower the central f_H2 and could produce or enhance the observed inside-out gradient. Because this concern directly targets the paper's main observational claim, please add a point-source + Sersic decomposition test or an equivalent color/AGN test, and quantify the allowed AGN fraction as a function of radius.","section":"Section 2.4 and Section 3.2"},{"comment":"The paper does not report the best-fit Sersic parameters (index n, effective radius, axis ratio) for the CO(1-0) and ALMA fits, even though the central result is that the molecular gas and SFR profiles are less centrally concentrated than the stellar profile. The F160W fit is quoted as n = 3.1 +/- 0.1 and Re = 2.6 +/- 0.1 kpc, but no analogous constraints are given for the lower-resolution data. This matters because the relative concentrations of the three components are load-bearing for the f_H2 and t_dep profiles. Please provide the fitted parameters and covariances (or a table or posterior summary) for all three components, and state explicitly whether the VLA CO source is resolved at a statistically significant level.","section":"Section 2.3 and Figure 4"},{"comment":"The t_dep radial gradient is presented in the abstract as a primary result, but the ALMA 343 GHz data are described in Section 2.3 as only 'marginally spatially resolved' and Section 3.1 concedes that the depletion-time trend has 'significantly larger uncertainties.' The text also concludes that the t_dep gradient is weaker evidence than the f_H2 gradient. Please either report a quantitative significance (for example, the posterior probability that the central t_dep is lower than the value at 2-4 kpc) or soften the abstract and conclusions to match the actual weight of evidence provided by the data.","section":"Section 3.1 and Figure 5"}],"minor_comments":[{"comment":"The statement about M/L gradients ('galaxies that are redder in their centers than the outskirts show falling M/L') is ambiguous; please specify whether M/L falls with increasing radius and clarify how that maps to the sign convention used in the text.","section":"Section 3.2"},{"comment":"The claim that 'the peak and subsequent decline at larger radii is real' should be backed by a quantitative significance estimate or by a statement of the signal-to-noise level of the CO detection at the radii of the f_H2 peak and decline.","section":"Section 3.1"},{"comment":"The color scale or color bar for the f_H2 map is not described in the text; consider adding a description of the color mapping and the masking threshold so that the figure is interpretable on its own.","section":"Figure 3"},{"comment":"The phrase 'the first time this phenomenon has been seen via observations of molecular gas in the high-redshift universe' could be softened to 'first resolved ground-state CO evidence' to avoid overclaiming, given the single-object nature of the study and the conversion-factor caveats discussed later in the paper.","section":"Abstract and Section 4"},{"comment":"The statement that 'This binning in velocity contains approximately 99% of the emission' would be clearer if the exact velocity width and the central velocity used for the binning were restated here, since the 125 km/s width is mentioned earlier but the verification of the 99% figure is not described.","section":"Section 2.1"}],"recommendation":"major_revision","confidential_remarks":"This is a well-executed single-object study with appropriate technical methods, but the central 'inside-out quenching' claim is vulnerable to AGN contamination of the F160W light, which is not addressed. The authors can likely mitigate this with existing data (e.g., a point-source plus Sersic fit to the HST image, or a color test for non-stellar central light) or at least with a bounding statement. The abstract overstates the depletion-time gradient relative to the ALMA resolution, and this should be reconciled. I therefore recommend major revision rather than rejection, provided the AGN test is added and the t_dep claims are recalibrated to the data quality."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead Spilker et al. 2019 on COSMOS 27289. The one thing you should know: this is the first spatially resolved ground-state CO(1-0) measurement of a compact z~2 star-forming galaxy, and it shows a central dip in molecular gas fraction and depletion time. If you care about how massive quiescent galaxies form, this is worth knowing even though it is a one-object case study.\n\nWhat I like: the analysis is careful and the methods are appropriate. Visibility-based Sersic fitting is the right way to handle marginally resolved interferometric data, and the authors propagate MCMC uncertainties into the radial profiles. The central suppression of f_H2 is not a fitted parameter; it emerges from direct ratio imaging under stated conversion assumptions. They are also honest about the fragility of the t_dep gradient and about the fact that plausible radial variations in M/L, alpha_CO, and dust temperature would strengthen the central deficit rather than erase it. The citation pattern is normal, and they directly address the apparent tension with Barro's high-J CO work.\n\nThe soft spots: it is one galaxy with peak CO S/N ~8 and only a few resolution elements. The t_dep profile is weak, and at large radii the profiles are dominated by Sersic extrapolation. The bigger issue is one the paper never addresses: a central AGN continuum in F160W would inflate the stellar mass at small radii and lower the central f_H2, producing or exaggerating exactly this pattern. The authors note that compact SFGs have a high AGN fraction, and the X-ray non-detection does not rule out an obscured AGN. They discuss stellar M/L gradients but not a point-source contribution. I would not call this disqualifying, but it is a real gap, and the inside-out quenching claim needs a point-source decomposition or an explicit argument why AGN contamination is negligible. The future-quenching forecast also depends on no substantial gas inflow in the next ~15 Myr; the authors flag this, and it is acceptable as a caveat.\n\nBottom line: this deserves a serious referee. I would send it to review with a request for an AGN-contamination test and a slightly more careful claim that the ground-state CO view is new, not the definitive proof of inside-out quenching. It is useful for observers and for anyone thinking about how compact SFGs transition to quiescence.","headline":"First resolved ground-state CO(1-0) view of a central gas deficit in a compact z~2 star-forming galaxy — a solid single-object result whose inside-out quenching narrative is plausible but still needs an AGN-contamination check in the stellar-mass map.","tokens_in":18882,"tokens_out":3865,"would_cite":true,"duration_ms":42850,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A massive z~2 galaxy is losing its molecular gas from the center outward, pointing to inside-out quenching.","keywords":["galaxy evolution","high-redshift galaxies","molecular gas","CO(1-0)","inside-out quenching","compact star-forming galaxies","star formation depletion","radio interferometry"],"falsifier":"A resolved map of a second molecular gas tracer, such as CO(2–1) or [CI], together with multi-band far-infrared data to measure the dust temperature gradient, could test the conversion-factor assumptions directly; if the radial profile of the depletion time becomes flat or rises toward the center after these corrections, the inside-out quenching claim would collapse. Alternatively, detecting gas inflow or a central concentration of dense gas on the ~10–15 Myr timescale would overturn the imminence of central quenching.","tokens_in":17776,"feed_emoji":"🔭","tokens_out":11064,"duration_ms":98474,"temperature":0.7,"pith_summary":"This paper seeks to show that COSMOS 27289, a massive compact star-forming galaxy at z=2.234, will cease star formation first in its center and later in its outskirts. High-resolution CO(1–0) imaging from the VLA reveals that both the molecular gas fraction $f_{\\mathrm{H_2}}=M_{\\mathrm{H_2}}/M_\\star$ and the depletion time $t_{\\mathrm{dep}}=M_{\\mathrm{H_2}}/\\mathrm{SFR}$ are lowest in the central 1–2 kpc and rise at radii of ~2–4 kpc. The authors interpret this as inside-out quenching: the center exhausts its gas in roughly 10–15 Myr while the outer disk lingers longer. If correct, it is the first molecular-gas evidence for inside-out quenching in the high-redshift universe and strengthens the case that compact star-forming galaxies are the immediate progenitors of z~2 quiescent galaxies.","feed_headline":"A z~2 galaxy's gas runs out from the center outward","feed_subtitle":"Resolved CO imaging shows the core's fuel will be gone in ~10-15 Myr, ahead of the outskirts.","key_machinery":"The load-bearing object is the resolved map of the CO(1–0) ground-state transition, observed with the VLA at ~0.25 arcsecond (~2 kpc) resolution and combined with lower-resolution C-array data to capture all the flux. From this map, plus HST F160W and ALMA 343 GHz imaging, the authors construct deconvolved radial surface brightness profiles of molecular gas, stellar mass, and obscured star formation, then form the ratios $f_{\\mathrm{H_2}}$ and $t_{\\mathrm{dep}}$. The argument's mechanism is a comparison of the shapes of these profiles: because the stellar mass is more centrally concentrated than the molecular gas, the gas fraction and depletion time are suppressed in the center, which is interpreted as imminent inside-out quenching. The analysis relies on a visibility-based Sérsic fitting method that avoids correlated uncertainties in sparsely resolved interferometric data, and on the galaxy's nearly face-on orientation, which makes the radial profiles relatively insensitive to inclination.","core_discovery":"The central discovery is that the radial distribution of molecular gas in COSMOS 27289 is not a scaled copy of the stellar mass distribution. Deconvolved radial profiles of CO(1–0) emission, HST F160W light, and ALMA 343 GHz continuum show that the molecular gas is the least centrally concentrated of the three tracers: the gas fraction rises from ~0.07 in the inner 1–2 kpc to roughly twice that value at 2–4 kpc before declining, and the depletion time rises from 10–15 Myr in the center by about a factor of two at 3 kpc. The authors argue that, given the galaxy's nearly face-on orientation and the likely radial trends in dust temperature, CO-to-$\\mathrm{H_2}$ conversion, and stellar mass-to-light ratio (which would all exaggerate the central suppression), this pattern implies that star formation will cease in the center before the outskirts—inside-out quenching. They further show good qualitative and reasonable quantitative agreement with a hydrodynamical simulation of a quenching galaxy, supporting the interpretation that rapid central gas consumption and outflows evacuate the core.","pith_inferences":["If the inside-out pattern is generic among compact star-forming galaxies, then integrated depletion times measured in unresolved surveys understate how quickly their centers are already running out of fuel; resolved gas imaging may be needed to locate the true quenching front.","The same visibility-based radial profile technique could be applied to a small sample of face-on compact SFGs at z~2 to test whether the central suppression of $t_{\\mathrm{dep}}$ is universal and whether the peak radius of the gas fraction tracks galaxy size or mass.","A direct observational test could come from a second epoch: after a rest-frame interval of ~10 Myr, a redshifted CO(1–0) or dust continuum observation might show the central surface brightness fading relative to the outskirts—a demanding but conceivable experiment.","The face-on orientation that makes COSMOS 27289 favorable also limits the sample; for inclined systems, kinematic deprojection of CO line maps would be needed to recover radial profiles, a natural extension of the method."],"forward_implications":["The galaxy's central kiloparsecs will exhaust their molecular fuel in ~10–15 Myr, so the central starburst should fade before the outer disk, producing a galaxy that looks quiescent in the center and still active outside.","Compact star-forming galaxies like COSMOS 27289 are plausible immediate progenitors of z~2 quiescent galaxies, consistent with their low gas fractions and short depletion times.","Radial variations in the physical conversion factors—lower central CO-to-$\\mathrm{H_2}$ conversion, warmer central dust, and a centrally concentrated stellar mass-to-light ratio—would only strengthen the observed central suppression, so the inside-out signal is a conservative lower limit.","The agreement with a hydrodynamical simulation of quenching suggests that central gas evacuation by rapid consumption and feedback is a generic feature at the end of a starburst, not an artifact unique to this object.","Ground-state CO(1–0) imaging, rather than higher-J CO transitions, is needed to trace the full molecular reservoir; higher-J lines would emphasize dense star-forming gas and could mask the central depletion."],"supporting_citations":[{"why":"Established the galaxy's low gas fraction and short depletion time from lower-resolution VLA CO(1–0) data, providing the target and baseline for this study.","marker":"Spilker et al. 2016a"},{"why":"Selected COSMOS 27289 as structurally similar to z~2 passive galaxies and noted its narrow line widths, supporting the face-on orientation argument.","marker":"van Dokkum et al. 2015"},{"why":"Supplied the residual-profile method used to measure deconvolved radial surface brightness profiles from the images.","marker":"Szomoru et al. 2010, 2012"},{"why":"Provides evidence that the CO-to-H2 conversion factor varies radially in nearby galaxies, supporting the claim that a single constant factor would exaggerate the central suppression.","marker":"Sandstrom et al. 2013"},{"why":"Standard calibration used to convert the far-infrared luminosity into the obscured star formation rate.","marker":"Kennicutt & Evans 2012"},{"why":"Analytic molecular fraction prescription applied to the simulation gas to estimate the molecular gas fraction for comparison.","marker":"Krumholz 2013"},{"why":"The hydrodynamical simulation of quenching whose radial gas fraction profile is compared quantitatively to the observed profile.","marker":"Tacchella et al. 2016"},{"why":"Earlier work interpreting rising specific SFR profiles as inside-out growth, providing the context for the molecular-gas result.","marker":"Tacchella et al. 2015"},{"why":"High-resolution ALMA imaging of compact SFGs showing concentrated SFR profiles, discussed as consistent with dissipative central buildup.","marker":"Barro et al. 2016"}],"fun_headline_variants":["Inside-out quenching seen in z~2 galaxy's gas","Galaxy core runs out of gas first at z~2","First molecular gas evidence for inside-out quenching","z~2 galaxy's gas dries up from the inside out"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The inside-out quenching forecast stands or falls on the assumptions that the conversion factors from observed light to stellar mass, molecular gas mass, and SFR do not vary radially in the opposite sense to the assumed constant values, and that no substantial gas flows into the center within the ~10–15 Myr depletion clock.","fun_headline_variants_meta":{"raw":{"variants":["Inside-out quenching seen in z~2 galaxy's gas","Galaxy core runs out of gas first at z~2","First molecular gas evidence for inside-out quenching","z~2 galaxy's gas dries up from the inside out"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000228,"raw_usage":{"total_tokens":1528,"prompt_tokens":1054,"completion_tokens":474,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":670,"completion_tokens_details":{"reasoning_tokens":407}},"tokens_in":670,"tokens_out":474,"duration_ms":5326,"temperature":1.0,"reasoning_tokens":407,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:49:50.692363+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A resolved map of a second molecular gas tracer, such as CO(2–1) or [CI], together with multi-band far-infrared data to measure the dust temperature gradient, could test the conversion-factor assumptions directly; if the radial profile of the depletion time becomes flat or rises toward the center after these corrections, the inside-out quenching claim would collapse. Alternatively, detecting gas inflow or a central concentration of dense gas on the ~10–15 Myr timescale would overturn the imminence of central quenching.","supporting_citations":[{"cited_title":"G., et al","cited_arxiv_id":null,"evidence_quote":"Supplied the residual-profile method used to measure deconvolved radial surface brightness profiles from the images."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Analytic molecular fraction prescription applied to the simulation gas to estimate the molecular gas fraction for comparison."},{"cited_title":"M., Dekel, A., et al","cited_arxiv_id":null,"evidence_quote":"High-resolution ALMA imaging of compact SFGs showing concentrated SFR profiles, discussed as consistent with dissipative central buildup."}],"review_version":1}