{"id":"3192bc5a-cd34-48c8-ac41-b3cf7901dbb0","arxiv_id":"2506.09152","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Using the NewHorizon and NewHorizon2 simulations, starless subhalos are traced to birth regions with low matter accretion, where gas cannot self-shield from UV background heating before reionization.","lead":"In computer simulations of Milky Way-like galaxies, most small dark matter clumps never form stars, and this empty fate is set by the region where they are born, not by later supernova explosions or tidal forces. The result offers a concrete explanation for the missing satellite problem: predicted dark matter clumps stay dark because their birth regions accrete matter too slowly to shield gas from the universe's first ultraviolet light.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim depends on the uniform, sharp reionization prescription: if radiative transfer changes which halos self-shield, the 'born starless' dichotomy could be an artifact.","rationale":"The paper is a careful simulation study: it uses high-resolution cosmological hydrodynamics from NewHorizon and NewHorizon2, explicitly checks for contamination, builds merger trees with strict criteria, and demonstrates that supernova feedback and ram-pressure/tidal stripping cannot transform a starred subhalo into a starless one because they remove gas but not stars. The reproduction of the Local Group satellite luminosity function is a real point of support. However, the paper's central claim is that the starless/starred dichotomy is set at birth by the accretion environment acting through reionization. That claim rests entirely on the treatment of reionization, which is a uniform UV background with a single self-shielding density threshold and no radiative transfer. The authors explicitly acknowledge this limitation in Section 4, and the cited RT studies (Zier et al. 2025, Ocvirk et al. 2020) show that ionization state in low-mass halos is sensitive to local sources and large-scale structure. Since the paper's entire causal narrative depends on which halos self-shield before reionization completes, the uniform-background approximation is the weakest load-bearing assumption. The proposed on/off test would directly determine whether reionization is the decisive agent or whether the birth-environment correlation is an independent accretion effect. If the latter, the phrase 'born to be starless' overclaims the role of reionization. The reader's weakest_assumption identifies exactly this point, so we agree. The paper is internally consistent and the main observational comparison is encouraging, so no change to the conditional verdict is warranted.","tokens_in":26173,"tokens_out":10903,"duration_ms":111265,"concrete_test":"Run a reionization on/off control: re-run NewHorizon2 (or a high-resolution subset containing the 26 MW analogs) with the UV background disabled. Re-identify the subhalos in the common mass range (10^8.4-10^8.9 Msun) and classify them as starred or starless using the same criteria. If a substantial fraction (e.g., >20%) of the 372 previously starless subhalos now form a resolved stellar population, reionization is the decisive agent and the birth-environment story holds. If they remain starless, the claim that reionization seals their fate is unsupported, and 'born to be starless' reduces to an accretion-rate effect independent of the reionization model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Starless subhalos are claimed to be 'born' because they reside in low-accretion birth regions, never reach the 0.01 H cm^-3 self-shielding density before the UV background turns on at z=10, and are subsequently heated. This causal chain is only as secure as the reionization prescription. The simulation adopts a spatially uniform Haardt-Madau background with an instantaneous analytic self-shielding correction (Eq. D6); there is no radiative transfer, no local ionizing sources, and no patchy or extended reionization. Real reionization is inhomogeneous and driven by low-mass galaxies, and the authors themselves write in Section 4 that the lack of full radiative transfer 'could modify our results' and cite Zier et al. (2025) finding differences that are 'much more severe in low-mass halos,' exactly the population at issue. If in a more realistic treatment some halos in underdense regions reionize later or are less exposed, they could retain gas and form stars, while halos in overdense regions could be ionized earlier, weakening or reversing the reported birth-environment correlation. Additionally, Eq. D6 as printed (nH,corr = nH / exp(-nH/0.01)) appears to be a sign error if the intended shielding factor is exp(-nH/0.01); as written, it would boost effective density and heating in dense gas, the opposite of shielding. This should be checked in the code, because if the literal formula is used, the central mechanism is even less reliable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the high-resolution cosmological simulations NewHorizon and NewHorizon2 to study why most subhalos around Milky Way analogs are starless. After showing that their simulated satellite counts match Local Group observations, the authors classify subhalos into starless and starred populations and test supernova feedback and infall-related environmental effects as possible causes of starlessness, finding neither able to transform a star-forming subhalo into a starless one. The paper argues instead that starless subhalos are born in low-accretion environments whose gas never reaches the self-shielding density before the z=10 UV background turns on, so the gas is heated and cannot cool to form stars. The conclusion is that the missing satellite problem is naturally alleviated by reionization physics and that starless subhalos are 'born to be starless, not made.'","tokens_in":26473,"tokens_out":5502,"duration_ms":55568,"significance":"If the causal claim holds, the paper offers a baryonic, reionization-based resolution of the missing satellite problem that does not require modifying dark matter and identifies the early accretion environment as the key predictor of present-day starlessness. The paper's strengths include a relatively large sample of 26 Milky Way analogs, high spatial resolution (34 pc in NewHorizon), a dense snapshot cadence (about 15 Myr), and a careful subhalo classification with merger-tree validation and reclassification of false starless/starred systems. The main claim is falsifiable in principle: it predicts a correlation between the large-scale accretion environment at early times and the satellite occupation fraction. However, the robustness of this result against the simplified reionization and self-shielding treatment is not yet established, and one printed equation appears to invert the self-shielding correction.","major_comments":[{"comment":"Equation (D6) as printed defines n_H,corr = n_H / exp(-n_H/(0.01 H cm^-3)), which equals n_H * exp(+n_H/(0.01 H cm^-3)). As written, this boosts the effective density of the densest gas, which is the opposite of self-shielding. If the simulation code uses the literal formula, the central mechanism described in Section 3.2.3 is inverted in dense gas; if the code instead implements n_H,corr = n_H * exp(-n_H/0.01), please correct Eq. (D6) and explicitly state the implemented form. This is load-bearing because the 'born to be starless' dichotomy depends on the self-shielding prescription.","section":"Appendix D, Eq. (D6)"},{"comment":"The central claim depends on the uniform UV background switched on at z=10 with an analytic self-shielding correction, rather than on radiative transfer with local ionizing sources. The manuscript itself notes in Section 4 the absence of radiative transfer and cites Zier et al. (2025) finding that differences are 'much more severe in low-mass halos,' which is exactly the population studied here. Because inhomogeneous or extended reionization can change which halos self-shield and when, the reported birth-environment dichotomy could be an artifact of the prescription. Please add robustness tests against plausible variations in the UV turn-on redshift, UV amplitude, or self-shielding threshold, or alternatively reframe the causal conclusion as conditional on the adopted reionization model.","section":"Section 4, reionization treatment"},{"comment":"It is not clear whether the comparison in Figures 8 and 9 uses the peak-mass-matched subsample defined in Section 3.2. The mass functions in Figure 5(a) show that starred and starless subhalos have different final mass distributions; if the birth-environment comparison uses all subhalos, the higher accretion rates of starred subhalos could simply reflect their higher masses rather than a distinct birth environment. Please state explicitly which sample is used and, if the full sample is used, repeat the analysis on the matched subsample to verify that the accretion-rate difference persists.","section":"Section 3.2.3, Figures 8 and 9"},{"comment":"The finding that starless subhalos never reach the self-shielding density is partly built into the model because the cooling tables impose a sharp transition at 0.01 H cm^-3 with a uniform UV background. The emergent part is which subhalos reach that density, and that part is interesting; however, the causal claim would be substantially stronger if the authors demonstrated that the correlation between early accretion and final starless fate persists when the self-shielding threshold or UV background model is varied within observationally allowed ranges.","section":"Section 3.2.3 and Appendix D"}],"minor_comments":[{"comment":"The star formation density thresholds are given as '5 (NH2) or 10 H cm^-3 (NH)', which is easy to misread; please write out '5 H cm^-3 for NewHorizon2 and 10 H cm^-3 for NewHorizon'.","section":"Section 2.1"},{"comment":"The background shading intended to indicate the reionization state is not defined in the caption or text; please add a legend or explicit description of the greyscale and the meaning of the epoch labels.","section":"Figure 8"},{"comment":"The term 'birthplace' is central to the argument but is used metaphorically; please define it operationally, for example as the position of the main progenitor at the first snapshot where it is identified in the merger tree.","section":"Sections 3.2.3 and 5"},{"comment":"The sentence 'We are particularly grateful to the referee for pointing us to numerous previous studies that were relevant to our investigation' is inappropriate for the published version and should be removed or rephrased.","section":"Acknowledgments"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a timely question and has solid simulation infrastructure, but the central causal claim is not yet robust to the simplified reionization treatment. The apparent sign error in Eq. (D6) should be checked in the code before publication, and the authors should clarify whether Figures 8-9 use mass-matched samples."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a careful simulation study that confirms the baryonic solution to the missing satellite problem and adds a quantitative data point: the starless subhalos were already on a separate track before reionization, because their birth regions accrete matter less efficiently. The authors are honest about novelty — Section 4 plainly states that the two key claims (reionization matters; birth environment and mass accretion history decide whether a subhalo stays dark) are not entirely novel, citing Sawala 2016, Simpson 2018, Benítez-Llambay 2019, Fitts 2017. The new bit is the resolved analysis of birth-environment accretion rates in a mass range where starred and starless subhalos overlap, plus a clean rule-out of supernova feedback and orbital stripping as primary causes. The sample handling is genuinely careful: peak-mass matching, merger-tree validation, reclassification of false starless and false starred subhalos, and satellite abundances match observed Local Group counts.\n\nThe soft spots are in the causal framing and the reionization modeling, not in the data processing. The \"born to be starless\" claim overstates what is, at bottom, a correlation between early accretion and later starless fate. There is no on/off reionization test or other controlled experiment that would nail causation; the authors acknowledge this. More importantly, the reionization prescription is a uniform UV background turned on at z=10 with an analytic self-shielding correction. No radiative transfer, no local ionizing sources, no patchy or extended reionization. The authors admit this could modify the results, and Zier et al. (2025) shows low-mass halos are exactly where the uniform-background approximation is most wrong. That is the load-bearing assumption.\n\nOne technical flag: Eq. D6 as printed, nH,corr = nH / exp(-nH/0.01), looks like a sign error if the intended shielding factor is exp(-nH/0.01). As written, it would boost effective density in dense gas, the opposite of shielding. If the code implements the literal formula, the central heating-suppression story is seriously compromised. A referee should ask for a code check. Also, no data are released, which limits reproducibility.\n\nThese issues are not disqualifying, but they do mean the central narrative should be softened from \"born\" to \"set early by environment under a particular reionization model.\" I would send this to review; the field needs high-resolution work like this, and the referee can ask for the Eq. D6 check and more careful causal language.","headline":"Careful, honest simulation work that confirms reionization suppresses faint satellites, but the 'born starless' framing runs ahead of the evidence and the reionization prescription is approximate.","tokens_in":27044,"tokens_out":3756,"would_cite":true,"duration_ms":35506,"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":"Starless subhalos are born, not made: their birth environments accrete too little matter to build self-shielding gas, so reionization heating prevents star formation before it begins.","keywords":["missing satellite problem","starless subhalos","reionization","self-shielding gas","cosmological simulations","dwarf galaxies","satellite galaxies","supernova feedback"],"falsifier":"Run the same subhalo census with radiative transfer of ionizing photons, or in a control run with the UV background removed or delayed, and compare the starless fraction among subhalos in the common mass range $10^{8.4}$–$10^{8.9}\\,M_\\odot$. The claim predicts that low-accretion subhalos would cool and form stars once UV heating is absent or patchy reionization lets more of them self-shield; if many low-accretion subhalos remain starless under those conditions, reionization heating cannot be the decisive cause of the starless population.","tokens_in":25961,"feed_emoji":"🌌","tokens_out":19833,"duration_ms":183480,"temperature":0.7,"pith_summary":"The long-standing 'missing satellite problem' is the gap between the many dark-matter subhalos that cosmological simulations predict around Milky Way-mass galaxies and the far fewer satellite galaxies actually seen. This paper argues that the gap closes through baryonic physics acting very early: subhalos born in regions where matter is accreted slowly never build gas dense enough to self-shield from the ultraviolet background that turns on at reionization, so their gas is heated and star formation never begins. Supernova feedback and stripping during orbital motion are shown to be secondary, since most starless subhalos never form stars at all and stripping removes gas but not stars. If the argument is right, a subhalo's observable fate is largely sealed before $z\\sim7$; no exotic dark matter or destructive feedback is required, and galaxy occupation below about $10^9\\,M_\\odot$ is set by assembly history and birth environment rather than final halo mass.","feed_headline":"Born starless: reionization seals subhalo fate at birth","feed_subtitle":"High-resolution simulations match observed satellite counts and trace the missing ones to birth environment, not feedback.","key_machinery":"The load-bearing mechanism is self-shielding of hydrogen gas against the uniform UV background that turns on at $z=10$. The simulations encode it analytically, via the corrected hydrogen density $n_{\\mathrm{H,corr}} = n_\\mathrm{H} / e^{n_\\mathrm{H}/(0.01\\ \\mathrm{H\\,cm^{-3}})}$, so that gas above the threshold $0.01\\ \\mathrm{H\\,cm^{-3}}$ stops being radiatively heated and can cool, while gas below it is heated and its net temperature change flips from cooling to heating. Whether a subhalo sits on the cooling or heating side of that threshold at reionization is traced back to the matter accretion rate measured in a 100 comoving kpc box centered on its birthplace, using merger trees built from stable member particles. The supporting machinery is the gravo-turbulent star-formation criterion, which forms stars only where local gravity overcomes thermal and turbulent pressure in cells above density thresholds of $10\\ \\mathrm{H\\,cm^{-3}}$ in NewHorizon and $5\\ \\mathrm{H\\,cm^{-3}}$ in NewHorizon2, plus a reclassification step that uses star-formation histories to remove interloper stars and separate 'true' from 'false' starred and starless subhalos.","core_discovery":"Across 26 Milky Way-analog systems drawn from the NewHorizon and NewHorizon2 cosmological zoom-in simulations, the cumulative abundance of satellite galaxies matches Local Group observations while the underlying subhalo population vastly outnumbers them: 2,032 starless subhalos against 416 starred ones. Among subhalos selected to have comparable peak masses, the two classes hold similar amounts of gas but differ sharply in cold gas: starless subhalos contain essentially none, so stars never form. The paper rules out the two standard explanations: supernova feedback depletes cold gas in starred subhalos only mildly, and 93.8% of starless subhalos never experience a supernova at all; ram-pressure and tidal stripping remove gas from infalling subhalos but leave pre-existing stars intact, so they cannot convert a starred subhalo into a starless one. The decisive difference is the birthplace: starless subhalos form where dark-matter and baryon accretion rates are lower, and their gas stays below the self-shielding density of $0.01\\ \\mathrm{H\\,cm^{-3}}$ as reionization completes at $z\\sim7$, so UV heating prevents cooling to the star-formation threshold of $5$–$10\\ \\mathrm{H\\,cm^{-3}}$. The paper's conclusion is in its title: starless subhalos are not made by feedback or stripping but born.","pith_inferences":["If birth environment is decisive, the ratio of luminous satellites to dark subhalos should vary with the large-scale environment: systems forming along dense, fast-accreting filaments should retain more luminous satellites than systems in slower regions, a trend testable with the growing census of satellite systems around Milky Way-mass hosts.","The mechanism chains reionization timing directly to the faint end of the galaxy luminosity function: an earlier or stronger UV background should push the mass scale at which half the subhalos go starless to higher masses, while delayed or patchy reionization should lower it.","Because starless subhalos are not empty but contain warm, pristine gas, they may be detectable in absorption or line emission despite emitting no starlight, which would turn an apparently unobservable population into a probe of reionization physics.","A no-reionization control run would separate the 'born' effect of slow accretion from the 'heated' effect of UV radiation; if a large starless population persists without any UV background, the mechanism would have to be rebalanced toward the accretion environment itself."],"forward_implications":["The classical missing satellite problem is resolved within standard cold dark matter by reionization acting on low-accretion birth environments, with no need for warm or self-interacting dark matter to suppress subhalo formation.","Because the starless/starred divide is set before reionization completes, a subhalo's luminous fate can in principle be predicted from its early merger-tree accretion history alone.","Supernova feedback and ram-pressure or tidal stripping regulate gas and quench star formation in already-starred subhalos, but neither mechanism can turn a starred subhalo into a starless one.","Below the overlapping mass range near $10^9\\,M_\\odot$, galaxy occupation is governed by assembly history and birth environment rather than final halo mass, so a sharp mass threshold for galaxy formation is the wrong description."],"supporting_citations":[{"why":"Supplies the NewHorizon simulation: its hydrodynamics, star formation and supernova feedback prescriptions, UV background, and self-shielding cooling tables on which the entire analysis runs.","marker":"Dubois et al. 2021"},{"why":"Supplies the NewHorizon2 twin simulation, whose 13 Milky Way analogs double the sample size.","marker":"Yi et al. 2024"},{"why":"Motivates the 0.01 H cm^-3 self-shielding threshold that separates UV-heated starless gas from cooling starred gas.","marker":"Rosdahl & Blaizot 2012"},{"why":"Provides the observed Milky Way and M31 satellite counts against which the simulated satellite abundance is compared.","marker":"O'Leary et al. 2023"},{"why":"Earlier hydrodynamical demonstration that baryonic heating and reionization reduce subhalo masses and satellite numbers; the result this work extends to the birth-environment mechanism.","marker":"Sawala et al. 2016"},{"why":"Supplies the gravo-turbulent star-formation criterion and the density thresholds that define when dense gas becomes stars.","marker":"Federrath & Klessen 2012"},{"why":"Provides the uniform UV background spectrum switched on at z=10 that heats the gas of starless subhalos.","marker":"Haardt & Madau 1996"},{"why":"Supplies the mechanical Type II supernova feedback scheme whose energy injection is tested and found not to be the main driver of starlessness.","marker":"Kimm et al. 2015"}],"fun_headline_variants":["Starless subhalos are born, not made","Birth environment, not feedback, creates starless halos","Reionization at birth dooms subhalos to starlessness","Subhalo star formation sealed by birth conditions","Missing satellites: born without stars, study finds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulations model reionization with a uniform UV background switched on at $z=10$ plus a shielding formula, rather than actually tracking ionizing radiation from individual sources, so the conclusion depends on that simplification correctly deciding which halos keep gas cool enough to form stars.","fun_headline_variants_meta":{"raw":{"variants":["Starless subhalos are born, not made","Birth environment, not feedback, creates starless halos","Reionization at birth dooms subhalos to starlessness","Subhalo star formation sealed by birth conditions","Missing satellites: born without stars, study finds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000224,"raw_usage":{"total_tokens":1534,"prompt_tokens":1092,"completion_tokens":442,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":708,"completion_tokens_details":{"reasoning_tokens":364}},"tokens_in":708,"tokens_out":442,"duration_ms":4959,"temperature":1.0,"reasoning_tokens":364,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:56:09.231178+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same subhalo census with radiative transfer of ionizing photons, or in a control run with the UV background removed or delayed, and compare the starless fraction among subhalos in the common mass range $10^{8.4}$–$10^{8.9}\\,M_\\odot$. The claim predicts that low-accretion subhalos would cool and form stars once UV heating is absent or patchy reionization lets more of them self-shield; if many low-accretion subhalos remain starless under those conditions, reionization heating cannot be the decisive cause of the starless population.","supporting_citations":[],"review_version":1}