{"id":"43cd5e81-7d0e-4192-bc0e-560a64871649","arxiv_id":"1909.00832","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Local photoionization feedback from a galaxy's own stars and hot gas reduces final stellar mass by about 20 percent and HI mass by about 50 percent in simulated galaxies spanning dwarf to massive spiral masses.","lead":"This paper simulates three galaxies with and without the extra heating and ionizing light from their own stars and hot gas. It finds that this 'local photoionization feedback' suppresses star formation by about 20 percent and roughly halves the neutral hydrogen mass by today.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 20% stellar-mass and ~50% HI claims rest on unswept fesc and density-only shielding; a two-point fesc sensitivity run would show whether the headline is robust.","rationale":"The reader's weakest assumption identifies exactly the fixed escape fraction in Eq. (8) and the possibility of significant line-of-sight absorption. My review concurs and extends the concern to the density-only shielding prescription and to the absence of any sensitivity analysis for the adopted fesc values. This is not an internal inconsistency: the paper is transparent about the approximation and lists future improvements. However, the central claim is quantitative, and the three headline numbers (20%, half, 0.95 dex) are all linear in the assumed local flux. A two-point parameter sweep is a cheap, decisive check that the paper has not performed. The small sample (one realization per galaxy) reinforces the concern but is secondary; it would matter even with perfect radiation physics. The reader's conditional verdict is appropriate, and my stress test does not move it.","tokens_in":33078,"tokens_out":6124,"duration_ms":70038,"concrete_test":"Re-run the Milky Way analogue g8.26e11 with FG09+LPF but with fesc,SFR set to 0.02 and to 0.1, all other parameters identical, and compare z=0 M*, MHI, and median inflow temperature to the fiducial run. If the 20% mass suppression and ~50% HI reduction persist within a few percent for both values, the escape-fraction choice is not load-bearing; if they move by more than ~5 percentage points, the quantitative claims require recalibration before acceptance. A complementary one-run swap of fesc,os from 1.0 to 0.5 would isolate the contribution of old stars, the dominant large-scale term in Eq. (9).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims in Section 6 (~20% M* suppression, ~50% HI reduction, 0.11-0.95 dex inflow temperature increase) are determined by the local photoionization field, which enters through Eqs. (8) and (9) with fesc,SFR fixed at 0.05 and fesc,os fixed at 1.0, and by the density-only attenuation exp(1-n/n0) for n > 0.1 cm^-3. These values are adopted from Kannan et al. (2014, 2016) and are not calibrated or varied in this paper. The authors explicitly state in Section 3 that a more physical attenuation would be column-density based (Sobolev-like) and that they plan to explore it in a future work. Because the claimed effects scale essentially linearly with the unattenuated LPF flux, a factor of 2 error in either escape fraction or in line-of-sight absorption would propagate directly into the photoheating rates and the equilibrium temperature that drives the mass suppression. With only one realization per galaxy, the apparent uniformity of the 20% suppression across three galaxies cannot be distinguished from stochasticity; the paper itself says a statistical sample is needed before observational tests. The headline numbers are therefore best read as model-dependent outcomes of a particular approximate treatment, not as a robust measurement of local radiation feedback.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper implements the optically thin local photoionization field (LPF) model of Kannan et al. (2014, 2016) in the SPH code GASOLINE2 and applies it to three zoom-in NIHAO galaxies spanning halo masses from ~10^11 to ~10^12.5 Msun. The model adds photoionization and photoheating from young star-forming regions, post-AGB stars, and hot gas on top of the FG09 UVB, with density-dependent attenuation. The authors compare each galaxy under HM12, FG09, and FG09+LPF and report that LPF suppresses z=0 stellar mass by ~20 per cent, reduces HI masses by roughly half, raises median CGM inflow temperatures by 0.11-0.95 dex without strongly changing inflow rates, reshapes HI discs and column-density distributions, and modestly increases bulge-to-total ratios. They also test UVB sensitivity by comparing HM12 and FG09.","tokens_in":33407,"tokens_out":5696,"duration_ms":55798,"significance":"The topic is timely: local radiation feedback is usually neglected in cosmological SPH simulations, and the study is one of only a few to apply the Kannan et al. model to galaxy-formation zoom-ins. The implementation on the gravity tree is computationally economical, the choice of three halo masses is useful, and the validation against the Milky Way ISRF at 8 kpc gives a welcome anchor. If the headline numbers are robust, the paper would support the view that ionizing radiation from local sources acts as a preemptive feedback that suppresses star formation and reshapes the CGM and HI structure of L* galaxies. The main caveats are the unvaried model parameters (fesc,SFR=0.05, n0=0.1 cm^-3, density-only attenuation) and the single realization per galaxy, which currently make the quantitative claims model-dependent rather than directly falsifiable.","major_comments":[{"comment":"The text and abstract report a ~20 per cent suppression of z=0 stellar mass, but the values in Table 2 imply only 11-13 per cent for all three galaxies (log M*/M⊙ = 8.98->8.93, 10.69->10.63, 11.29->11.23 for FG09 vs FG09+LPF). If the 20 per cent figure refers to a different epoch or an averaged quantity, this should be stated explicitly; as written, the headline number is inconsistent with the table.","section":"§5.1, Table 2"},{"comment":"The central quantitative results scale linearly with the adopted escape fractions (fesc,SFR=0.05, fesc,os=1.0) and with the density-only shielding exp(1-n/n0); none of these parameters is varied or calibrated here. The paper itself notes that a Sobolev-like column-density attenuation would be more physical. Because a factor of two change in these parameters directly changes photoheating and hence the equilibrium temperature that drives the claimed 20 per cent mass suppression and inflow-temperature increases, the headline claims need a sensitivity test (e.g., fesc,SFR = 0.025 and 0.1, or a column-based attenuation) before they can be presented as more than conditional outcomes of this particular parameter choice.","section":"§3, Eqs. (8)-(9) and attenuation after Eq. (14)"},{"comment":"Each galaxy is represented by a single realization, and stochastic variations in star formation and feedback are large (the SNe bubble morphology in Figure 8 is explicitly stochastic). The apparent uniformity of the mass suppression across three galaxies cannot be distinguished from run-to-run scatter without multiple initial conditions. The paper acknowledges the need for a statistical sample, but the central claims would be strengthened by at least one repeat run or by explicit bootstrap or error estimates.","section":"§5.1, §5.2, Figures 3 and 4, Table 2"}],"minor_comments":[{"comment":"There are several typos: 'halfed' should be 'halved', 'attentuated' should be 'attenuated', 'resonable' should be 'reasonable', and 'cuves' in the Figure 4 caption should be 'curves'.","section":"Abstract and Section 5.1"},{"comment":"The text says 'see Section 11' when referring to the shielding discussion; this should be 'see Section 3'.","section":"§5.2.2, paragraph before Figure 11"},{"comment":"The statement that Γ(HI) ∝ R^-2 is a 'prediction' should be softened, since it follows algebraically from the inverse-square assumption in Eq. (8) and the authors already note this in the text.","section":"§5 and Figure 2"},{"comment":"The text alternates between 'five types' and 'three types' of local radiation sources; since the hot gas is split into three temperature bins, a consistent wording such as 'three source types, five spectral components' would remove ambiguity.","section":"§3"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the paper's advertised 20 per cent stellar-mass suppression is contradicted by its own Table 2, which gives 11-13 per cent; this needs correction before publication. The larger scientific issue is the absence of sensitivity runs for fesc,SFR and the attenuation prescription, which are load-bearing for the quantitative conclusions. The manuscript fits the journal's scope and, with these revisions, could be acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look. The genuinely new piece is the GASOLINE2 implementation of the Kannan et al. LPF model and its application to three NIHAO galaxies spanning dwarf to massive spiral. The mass-dependent inflow temperature increase (0.11 dex to 0.95 dex) is new and is the most interesting result; the HI column density shifts are also a useful prediction. The 20% stellar-mass suppression and ~50% HI reduction are plausible and consistent with the earlier Kannan et al. work, but they are not robust measurements — they are outcomes of one approximate treatment.\n\nThe paper is well written, the code description is clear, the comparison with Hopkins et al. and Kannan et al. is honest, and the authors flag their own limitations (density-only attenuation, fixed escape fractions, need for a statistical sample). The ISRF comparison for the MW analogue is a nice sanity check.\n\nSoft spots: three galaxies, one realization each, so the uniformity of the 20% suppression could be partly stochastic. The escape fractions are fixed at fesc,SFR=0.05 and fesc,os=1.0 with no sensitivity runs; the attenuation exp(1-n/n0) is a smooth ad hoc version of the step function in Kannan et al. The authors acknowledge a column-density-based Sobolev attenuation would be better. The UVB comparison is confounded because the HM12 and FG09 runs use different metal cooling tables, so part of the UVB difference may be table differences rather than the UVB itself. The R^-2 Gamma(HI) profile is an algebraic consequence of the inverse-square law, not a new physical prediction.\n\nThose soft spots are real but not fatal. The paper's conclusions are appropriately cautious and it explicitly defers observational tests to a statistical sample. I'd send it to a serious referee. Ask for a two-point fesc sensitivity run (say fesc=0.1) and a clearer statement of the cooling-table consistency in the UVB comparison, plus ideally one repeat realization per galaxy to check stochasticity.\n\nWho's it for: people running galaxy formation simulations with SPH or mesh codes, and observers working on HI and CGM absorption. It deserves a referee.","headline":"A competent, honest application of an existing LPF model to three NIHAO galaxies; the mass-dependent inflow temperature result is new, but the headline 20% M* and 50% HI numbers need a sensitivity run before they are robust.","tokens_in":34067,"tokens_out":2382,"would_cite":true,"duration_ms":26507,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that a galaxy's own ionizing radiation, modeled in an optically thin approximation, suppresses stellar mass growth by about 20 per cent by z=0 and roughly halves HI mass in simulated dwarf, Milky Way-like, and massive…","keywords":["galaxy formation","local photoionization feedback","cosmological hydrodynamical simulations","circumgalactic medium","HI content","ultraviolet background","optically thin approximation","GASOLINE2"],"falsifier":"Measure the escape fraction of ionizing radiation from a Milky Way-mass galaxy at low redshift; a value several times higher or lower than 0.05 would shift the predicted stellar-mass suppression and inflow-temperature increase outside the paper's quoted ranges. Alternatively, search for Lyman-$\\alpha$ absorption from ~$10^{4}$ K gas at 0.1–0.4 virial radii around massive spirals at z~0: the LPF model predicts almost none, so a detection of such gas would rule out the strong-heating picture.","tokens_in":32905,"feed_emoji":"🌌","tokens_out":7761,"duration_ms":78905,"temperature":0.7,"pith_summary":"This paper asks what happens to galaxies when the ionizing radiation produced by a galaxy's own stars and hot gas is added on top of the standard uniform ultraviolet background in cosmological simulations. It reports that this 'local photoionization feedback' acts as a preemptive brake on star formation: in three simulated galaxies spanning dwarf to massive-spiral mass, the final stellar mass is reduced by about 20 per cent relative to runs with only the uniform background, and the neutral hydrogen (HI) mass is roughly halved. The field also heats the gas flowing into galaxies, raising the median inflow temperature by 0.11 dex in the dwarf and 0.95 dex in the massive spiral, while leaving the inflow rate nearly unchanged. A sympathetic reader would care because most galaxy simulations omit this local field, and the predicted changes affect observable quantities such as HI content, stellar disc structure, and the temperature of gas in and around galaxies.","feed_headline":"A galaxy's own starlight slashes its stellar mass by 20 percent","feed_subtitle":"Simulations show the local ionizing field also halves HI content and heats inflowing gas by up to an order of magnitude.","key_machinery":"The load-bearing machinery is the local photoionization field (LPF) computation added to the GASOLINE2 SPH code. Five source terms are summed at each gas particle: young stars and supernova remnants (with a 5 per cent escape fraction), post-AGB stars (100 per cent escape fraction), and hot gas in three temperature bins around $10^{6}$, $10^{7}$ and $10^{8}$ K for Bremsstrahlung emission. Each contribution scales as the inverse square of the distance to the source, and an exponential attenuation factor exp(1-n/n0) with n0=0.1 $cm^{-3}$ shields the densest gas. The five fields are evaluated on the gravity tree at negligible extra cost, and enter the photoionization and photoheating rates of hydrogen, helium, and metals alongside the chosen ultraviolet background. This machinery converts a spatially uniform radiation assumption into an inhomogeneous one, and it is the heating it deposits in the halo gas that drives the paper's results.","core_discovery":"The paper's central claim is that an optically thin treatment of local ionizing sources—young massive stars and supernova remnants, post-AGB stars, and X-ray emitting hot gas—changes the thermal state of the gas that would otherwise cool onto galaxies, and thereby changes the galaxies that form. The specific findings are that at z=0 the local radiation field suppresses stellar mass growth by ~20 per cent for all three galaxies, reduces HI mass by roughly a factor of two, raises the median temperature of inflowing gas from 0.11 dex (dwarf) to 0.95 dex (massive spiral) without significantly altering the inflow rate, and shifts the galaxy/CGM boundary inward by raising the equilibrium temperature at that interface. As a result the HI discs become thinner and less extended, the fraction of cold star-forming gas drops by at least half, and the distribution of HI column densities shifts to lower values. The choice between two ultraviolet background models matters mainly for the dwarf galaxy, where the stronger background mimics a weaker background plus the local field.","pith_inferences":["If the true escape fraction of ionizing photons from star-forming regions differs substantially from the assumed 0.05, all the quoted magnitudes scale roughly linearly; an order-of-magnitude change would push the 20 per cent suppression and 0.95 dex heating outside their quoted ranges. This is a testable calibration rather than a separate prediction.","Because the optically thin approximation ignores line-of-sight absorption, the model likely overestimates the radiation field in shadowed dense regions and underestimates it where gas is illuminated from many directions; full radiative-transfer runs on the same three galaxies would bracket the error.","The result that inflow rates stay nearly constant while inflow temperatures rise suggests that this feedback works by raising the cooling time rather than by throttling accretion; a similar dichotomy may help interpret 'preventive' versus 'ejective' feedback in other simulations."],"forward_implications":["Galaxy simulations that include only a uniform ultraviolet background likely overproduce stellar mass by ~20 per cent and HI mass by up to a factor of two relative to runs that include the galaxy's own radiation field.","The local field's main effect is to warm the circumgalactic gas and move the boundary between cold disc gas and hot halo inward, implying less cold gas in the halo and thinner, smaller HI discs.","For galaxies near and above the Milky Way mass, the choice of ultraviolet background model has little effect on stellar mass, so the local field, not the background, is what matters there.","The predicted lack of ~10^4 K gas in the haloes of massive spirals at 0.1–0.4 rvir provides a direct observational target for Lyman-alpha absorption studies."],"supporting_citations":[{"why":"Provides the original optically thin local radiation field implementation in GASOLINE, including the escape fractions and SEDs for young stars and post-AGB stars.","marker":"Kannan et al. (2014)"},{"why":"Extends the model to include hot gas Bremsstrahlung and produces the CLOUDY cooling table used for metal heating and cooling in the LPF model.","marker":"Kannan et al. (2016)"},{"why":"Supplies the NIHAO galaxy sample and the standard HM12 simulations that this paper resimulates with the new feedback.","marker":"Wang et al. (2015)"},{"why":"Provides the FG09 ultraviolet background on top of which the local photoionization field is added in the LPF runs.","marker":"Faucher-Giguère et al. (2009)"},{"why":"Provides the HM12 ultraviolet background used in the standard NIHAO runs and as the comparison UVB model.","marker":"Haardt & Madau (2012)"},{"why":"Defines the non-equilibrium cooling and heating framework in GASOLINE that the local photoionization field modifies.","marker":"Shen et al. (2010)"},{"why":"Describes GASOLINE2, the code in which the LPF machinery is implemented and all simulations are run.","marker":"Wadsley et al. (2017)"},{"why":"Serves as the contrasting study of local radiation feedback using ray-based and moments-based schemes, providing the comparison for stellar-mass effects.","marker":"Hopkins et al. (2018a)"}],"fun_headline_variants":["Galaxy's own starlight slashes stellar mass by 20% and halves HI","Local ionizing radiation suppresses galaxy stellar growth by a fifth","A galaxy's internal starlight heats inflowing gas and trims its mass","Feedback from a galaxy's own stars cuts HI gas in half and shrinks mass","Starlight within a galaxy reduces its stellar mass by 20% and HI by half"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results rest on the assumption that a 5 per cent escape fraction for star-forming radiation plus an optically thin inverse-square falloff represents how ionizing photons actually leave star-forming regions and cross the halo, because any substantial change in that escape fraction would change the size of the claimed heating and stellar-mass suppression.","fun_headline_variants_meta":{"raw":{"variants":["Galaxy's own starlight slashes stellar mass by 20% and halves HI","Local ionizing radiation suppresses galaxy stellar growth by a fifth","A galaxy's internal starlight heats inflowing gas and trims its mass","Feedback from a galaxy's own stars cuts HI gas in half and shrinks mass","Starlight within a galaxy reduces its stellar mass by 20% and HI by half"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001345,"raw_usage":{"total_tokens":5512,"prompt_tokens":1041,"completion_tokens":4471,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":657,"completion_tokens_details":{"reasoning_tokens":4366}},"tokens_in":657,"tokens_out":4471,"duration_ms":34298,"temperature":1.0,"reasoning_tokens":4366,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:34:50.739647+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the escape fraction of ionizing radiation from a Milky Way-mass galaxy at low redshift; a value several times higher or lower than 0.05 would shift the predicted stellar-mass suppression and inflow-temperature increase outside the paper's quoted ranges. Alternatively, search for Lyman-$\\alpha$ absorption from ~$10^{4}$ K gas at 0.1–0.4 virial radii around massive spirals at z~0: the LPF model predicts almost none, so a detection of such gas would rule out the strong-heating picture.","supporting_citations":[{"cited_title":"S., Macci \\`o A","cited_arxiv_id":null,"evidence_quote":"Provides the original optically thin local radiation field implementation in GASOLINE, including the escape fractions and SEDs for young stars and post-AGB stars."}],"review_version":1}