{"id":"14b33ce5-d06f-4fa9-9f3f-2b2c40bc69e0","arxiv_id":"2507.13904","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":12,"one_line_summary":"N-body simulations of the Sparkler clumps predict that survivors keep half-mass radii of 20-50 pc to z=0, making them too large to evolve into present-day globular clusters.","lead":"This paper uses N-body simulations to track the 9.23 Gyr dynamical evolution of the stellar clumps around the z=1.4 Sparkler galaxy. Its main finding is that the clumps that survive outside the galaxy remain roughly ten times too large to be present-day globular cluster progenitors, while the most massive clumps spiral into the center.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Final clump sizes may be set by unbound tidal debris rather than the bound remnant, so the 'too large' conclusion needs a bound-mass re-analysis.","rationale":"The reader's weakest assumption is observational: the clump effective radii might be overestimated by JWST resolution or lensing magnification, in which case the simulated sizes would inherit inflated initial values. That is a legitimate concern, and Sect. 6.1 already discusses it. My stress-test identifies a different, purely numerical weak spot: the final sizes are measured from a Plummer fit to all particles in a fixed aperture, with no binding-energy selection. This matters most for the low-mass survivors in the disk-shock runs, where a large fraction of the enclosed mass is expected to be tidally stripped debris. If the bound remnant is compact, the central claim is not established; if it remains extended, the central claim is robust. The proposed test is one specific re-analysis of existing snapshots, so it can settle the issue without new observations. I therefore keep the reader's CONDITIONAL verdict unchanged: the paper is valuable and internally coherent, but this size measurement should be verified before the abstract's headline conclusion is taken as final.","tokens_in":31366,"tokens_out":12494,"duration_ms":180742,"concrete_test":"Take the final snapshots of the Sect. 5.2 disk-shock runs (and, for control, the spherical-host runs) and compute the bound remnant of each clump by iteratively removing particles with positive total energy in the clump rest frame, using the shrinking-sphere centre from Appendix A, until the bound mass converges. Then compute the 3D and projected half-mass radii of the bound particles only. If the bound half-mass radii for clumps with bound mass >10^5 Msun fall below 10 pc, the 'too large' conclusion fails; if they remain above 20 pc for bound masses >10^6 Msun, the central claim survives this check.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The size conclusion in Sect. 6.1 is built on Plummer fits to the final snapshots of the tidal-stripping runs (Appendix A). The fit includes all particles within 4rs,i of the clump centre, with no selection of gravitationally bound particles. The disk-shock runs (Sect. 5.2) are specifically designed to generate tidal tails (Fig. 11), and for low-mass clumps the retained mass within 4rs,i falls to roughly 15-40% (Fig. 10, bottom). In such a two-component system - a compact bound core plus an extended tail of unbound debris - a single-Plummer fit to the total density profile can return a scale radius set by the tail rather than by the object that would survive as a GC. If, after removing unbound particles, the bound remnants of the low-mass survivors have half-mass radii below about 10 pc while retaining more than ~10^5 Msun, then the abstract's claim that 'the remaining clumps are too large in size to be progenitors of GCs' does not follow from these simulations. This can be checked directly from the snapshots and is independent of the observational magnification assumption flagged in Sect. 6.1.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies whether the stellar clumps observed around the z=1.4 lensed Sparkler galaxy can evolve into globular clusters by z=0. The authors build initial conditions from the observed clump masses, effective radii, and distances, then run two complementary suites of N-body simulations with the AREPO code: (i) dynamical-friction runs in which point-mass clumps orbit inside a live NFW/Hernquist halo, and (ii) tidal-stripping runs in which resolved Plummer clumps orbit in static spherical or disk+halo potentials. They find that dynamical friction sends most clumps more massive than about 10^7 M_sun to the central regions, that tidal shocks from a disk can disrupt or heavily strip low-mass clumps, and that the surviving clumps retain half-mass radii of roughly 20-50 pc. Combining the two processes with a semi-analytic mass-loss model shifts the survivor mass peak to about 2e6 M_sun, consistent with very massive globular clusters, but the authors conclude that the surviving clumps are too large in size to be globular-cluster progenitors.","tokens_in":31709,"tokens_out":13083,"duration_ms":162045,"significance":"If the size conclusion survives detailed scrutiny, the paper is significant: it directly tests a JWST-motivated hypothesis that the Sparkler clumps are proto-globular clusters, using explicit N-body integrations rather than purely analytic estimates. The paper also has genuine strengths: the resolved clumps are followed with 10^5 particles, three spatial configurations are explored, the eccentric-orbit case is treated in an appendix, and the assumptions that maximize tidal effects are stated transparently. The headline result is falsifiable and observationally meaningful. However, the central claim currently rests on a size measurement that may be contaminated by unbound tidal debris, and the quantitative mass claim depends on an extrapolation of a maximum-stripping calibration; both points need to be addressed before the conclusion can be accepted.","major_comments":[{"comment":"The final scale radii rs,f are obtained by fitting a single Plummer profile to the total density profile of all particles within 4 rs,i, with no selection of gravitationally bound particles. The disk-shock runs (Sec. 5.2, Fig. 11) generate extended tidal tails, and for low-mass clumps the retained mass within 4 rs,i can be only a small fraction of the initial mass (bottom panel of Fig. 10). In such a two-component system, a single Plummer fit measures the envelope containing both the bound remnant and the unbound debris, so the fitted rs,f is not necessarily the size of the object that would survive as a globular cluster. Because the abstract's central claim that 'the remaining clumps are too large in size to be progenitors of GCs' is based directly on Fig. 12, the authors should repeat the analysis on the bound remnant, using an iterative energy-based unbinding procedure, and report the bound mass and bound half-mass radius for each run in Secs. 5.1 and 5.2. Without this re-analysis, the size conclusion is not established for the low-mass survivors.","section":"Sec. 6.1, Fig. 12, and Appendix A"},{"comment":"The tidal mass-loss relation of Eq. (17) is calibrated on resolved-clump runs at r0 = 1 kpc and inclination theta = pi/6, parameters chosen explicitly to maximize disk shocks, and is then applied without any radial dependence to all surviving clumps in the spherical dynamical-friction runs, which end at distances ranging from about 1 to 10 kpc (Figs. 6 and 8). The predicted peak near 2e6 M_sun is therefore conditional on a maximum-stripping assumption, and the statement that the corrected mass distribution is 'compatible with massive GCs' is a lower limit on the true final masses. The authors note the maximization at the end of Sec. 6.2, but they do not quantify how the peak and the fraction of over-massive survivors would change if the disk-shock model were applied only to clumps with pericentres close to 1 kpc, or if a distance-dependent calibration were used. This quantification is needed to support the mass claim in the abstract.","section":"Sec. 6.2, Eq. (17), and Fig. 13"},{"comment":"The size-mass relation used to set the initial Plummer scale radii has a poorly constrained intercept, q = 0.19 +/- 0.65, and is fitted to only seven resolved clumps. The simulations then extrapolate this relation to masses as low as 1e6 M_sun and as high as 6e7 M_sun, extending beyond the observed range 10^6.3-10^7.4 M_sun. Because the final sizes in Fig. 12 are directly tied to these initial radii, the authors should propagate the 1-sigma (or 2-sigma) uncertainties in m and q into the final-size predictions, or restrict the size conclusion to the mass range directly sampled by the resolved clumps.","section":"Sec. 2.1, Eq. (1), and Fig. 12"}],"minor_comments":[{"comment":"The text says 'For each spatial configuration we ran 15 simulations', but Sec. 4.4.1 and the captions of Figs. 6 and 7 refer to 'the 30 simulations'. This inconsistency should be corrected, since it affects the interpretation of the quoted survival fractions and their uncertainties.","section":"Sec. 4.3 and Figs. 6-7"},{"comment":"The caption refers to 'the size-mass relation of Eq. 2.1' but the correct reference is Eq. (1).","section":"Fig. 12 caption"},{"comment":"The third row is labelled 'MN2', duplicating the second row; it should presumably be labelled 'MN3'.","section":"Table 5"},{"comment":"The bottom-panel caption says 'same as the left panel', but it should say 'same as the top panel'.","section":"Fig. 10 caption"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is suitable for A&A in scope and method. The main risk is that the headline negative claim may be an artefact of fitting the total (bound plus unbound) density profile rather than the bound remnant; this is a fixable but load-bearing issue. I have no concerns about citation practices or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is the first time anyone has put the Sparkler clumps through actual dynamical evolution, and the headline result is conditional but plausible — most of the clumps either sink to the centre by z=0 or survive as objects that are still too big to be globular-cluster progenitors. The paper is worth reading and worth refereeing.\n\nWhat it does well: the authors use standard tools (live halo for dynamical friction, resolved Plummer clumps in an external potential for tidal stripping, Chandrasekhar drag for orbit calibration) and apply them carefully to one specific lensed galaxy. They test three spatial configurations because the morphology is unconstrained, and they are upfront about the assumptions. The survival fractions (~40–60%) and the prediction that clumps above ~1e7 Msun sink are sensible. The comparison with bulge fossil fragments and NSCs is useful.\n\nThe soft spot, and it is a real one: the claim that surviving clumps stay 20–50 pc relies on Plummer fits to the total density within 4rs,i, with no selection of bound particles. In the disk-shock runs the low-mass clumps keep only 15–40% of their mass inside that radius and have obvious tidal tails (Fig. 11). A single Plummer fitted to the total profile can return a scale radius set by the debris rather than by the compact bound remnant. If the bound remnants have half-mass radii below ~10 pc and still hold ~1e5 Msun, the abstract's 'too large' conclusion does not follow. This is checkable from their snapshots, and the authors should do that before publication.\n\nThe second issue is that the size–mass relation used to initialise clumps is fitted to only seven resolved clumps, with intercept q = 0.19 ± 0.65. The paper does not propagate that uncertainty. A lower intercept would mean more compact initial conditions and, presumably, more compact survivors. The qualitative conclusion about sizes is therefore more fragile than the paper lets on.\n\nThe mass-distribution shift to ~2e6 Msun in Section 6.2 uses a semi-analytic combination with no propagated uncertainties. That is a lesser concern; the qualitative shift is believable.\n\nBottom line: this is a solid case study for anyone interested in high-z star clusters or GC formation. It deserves peer review. I would ask the referee to focus on the bound-mass re-analysis of the final sizes and on the sensitivity to the size–mass intercept. If those hold up, the paper is publishable as a valuable addition to the Sparkler literature.","headline":"A careful, useful dynamical-evolution study of the Sparkler clumps, but the headline size conclusion needs a bound-mass re-analysis before it is solid.","tokens_in":32208,"tokens_out":5884,"would_cite":true,"duration_ms":72695,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that the stellar clumps around the z≈1.4 Sparkler galaxy that survive to z=0 remain 20–50 pc in half-mass radius, about ten times the size of local globular clusters, and are therefore too large in size to be…","keywords":["globular cluster formation","Sparkler galaxy","stellar clumps","dynamical friction","tidal stripping","N-body simulations","gravitational lensing","JWST"],"falsifier":"Re-estimate each clump's lensing magnification and intrinsic size from a deeper lens model or from resolved stellar kinematics: if the true half-mass radii come out near a few parsecs rather than 30–50 pc, the size-based conclusion fails. An independent check is to search local galaxies of roughly $10^{10}$ solar masses for surviving extended clusters with radii of 20–50 pc and masses above $10^6$ solar masses; their absence would support the paper's prediction.","tokens_in":31165,"feed_emoji":"🔭","tokens_out":9436,"duration_ms":96564,"temperature":0.7,"pith_summary":"The paper asks whether the bright stellar clumps seen around the $z\\approx1.4$ Sparkler galaxy—objects with globular-cluster-like ages and metallicities but masses and sizes about ten times larger than local globular clusters—can evolve into globular clusters by $z=0$. Using N-body simulations of dynamical friction and tidal stripping over 9.23 Gyr, the authors find that clumps more massive than about $10^7\\,\\mathrm{M_\\odot}$ sink into the galactic centre, while the clumps that survive beyond 1 kpc lose enough mass, through tidal shocks from a stellar disk, to match massive globular clusters in mass. Their sizes, however, stay between 20 and 50 pc, about ten times larger than present-day globular clusters. The paper's central conclusion is that the surviving Sparkler clumps are too large in size to be globular-cluster progenitors, and that most of the clumps are instead destined for the bulge or a nuclear star cluster.","feed_headline":"Surviving Sparkler clumps are 10x too big for globular clusters","feed_subtitle":"N-body runs from z=1.4 to today show survivors keep 20–50 pc radii even after tidal stripping.","key_machinery":"Two complementary N-body setups carry the argument. In the first, each clump is a single particle moving in a live halo of $10^7$ dark-matter particles built to match a Sparkler-like NFW/Hernquist potential, so dynamical friction arises self-consistently; the clump system is initialised in three geometries—face-on disk, edge-on disk, and spherical—to bracket the unknown morphology. In the second, clumps are resolved as Plummer spheres of $10^5$ particles orbiting in a static external potential that includes the same halo and, in one set of runs, an exponential stellar disk (represented by three Miyamoto-Nagai disks fixed at $z=0$ mass and size) whose crossings produce tidal shocks. The two channels are joined by fitting the simulated mass loss as a function of initial mass and applying it before or after the dynamical-friction selection, bracketing the order in which the processes act.","core_discovery":"For a Sparkler-like galaxy with a dark-matter halo reaching $M_{200}\\simeq5\\times10^{11}\\,\\mathrm{M_\\odot}$ by $z=0$ (the high end of the inferred range), the simulations predict that only about 40–60% of the ten extraplanar clumps survive outside 1 kpc. Clumps with stellar masses above roughly $10^7\\,\\mathrm{M_\\odot}$ are dragged by dynamical friction into the central regions on timescales shorter than the 9.23 Gyr available. Without tidal stripping the survivor mass distribution peaks near $5\\times10^6\\,\\mathrm{M_\\odot}$, implying unusually over-massive clumps at $z=0$; when mass loss from disk tidal shocks is included and corrected for, the peak shifts to about $2\\times10^6\\,\\mathrm{M_\\odot}$, consistent with the most massive globular clusters known. Yet the same simulations show that the surviving clumps keep half-mass radii of 20–50 pc down to $z=0$. The paper concludes that most of the Sparkler clumps end up as bulge fossil fragments or contributors to a nuclear star cluster, while those that remain outside are too large in size to be the progenitors of today's globular clusters.","pith_inferences":["A direct corollary of the paper's bracketing is that the order in which tidal stripping and dynamical friction act changes the fraction of survivors (from about 0.42 to 0.60) but leaves the final mass distribution nearly unchanged, which may make the relative timing hard to constrain from masses alone.","If the intrinsic sizes really are 20–50 pc, one would expect to find a population of extended, massive clusters—objects unlike ordinary globular clusters—in local galaxies of roughly $10^{10}\\,\\mathrm{M_\\odot}$; searching nearby analogues for such extended clusters is a testable consequence the paper does not pursue.","The paper assumes no dark-matter subhalo around the clumps; if they were embedded in mini-halos, its own dynamical-friction results imply they would sink even faster, which would suppress the over-massive survivors rather than solve the size problem.","A revision of the lens model that shrinks the inferred radii to a few parsecs would not by itself rescue the globular-cluster scenario: the paper's argument shows the clumps would then be dense enough to resist tidal stripping and would survive as over-massive outliers."],"forward_implications":["Clumps more massive than about $10^7\\,\\mathrm{M_\\odot}$ should sink into the Sparkler's central regions by $z=0$, where they can feed bulge growth or merge into a nuclear star cluster.","Only about 40–60% of the observed extraplanar clumps are expected to survive beyond 1 kpc, so the clump population around the galaxy thins substantially with time.","Correcting survivors for tidal stripping shifts their peak mass from about $5\\times10^6$ to $2\\times10^6\\,\\mathrm{M_\\odot}$, placing the descendants at the high-mass end of the globular-cluster mass function.","Even after tidal stripping, survivor half-mass radii remain 20–50 pc, about ten times larger than local globular clusters, so mass-compatible descendants are still not size-compatible globular clusters.","If the lensing magnification were underestimated by about a factor of 10, the clump masses would become consistent with massive globular clusters while their sizes would remain up to five times too large."],"supporting_citations":[{"why":"discovered the Sparkler galaxy and its system of stellar clumps, defining the observational target.","marker":"Mowla et al. 2022"},{"why":"supplies the ten clump masses, effective radii, and source-plane distances that seed and calibrate the simulations.","marker":"Claeyssens et al. 2023"},{"why":"classifies the sub-sample as globular-cluster candidates from ages and metallicities and provides the galaxy's stellar mass.","marker":"Adamo et al. 2023"},{"why":"establishes that the clumps' projected distances resemble the disk/halo globular-cluster distribution, motivating the comparison.","marker":"Forbes & Romanowsky 2023"},{"why":"provides the stellar-to-halo mass relation used to fix the Sparkler halo mass and its growth to z=0.","marker":"Behroozi et al. 2019"},{"why":"supplies the dynamical-friction timescale and tidal-radius estimates used to frame the simulation outcomes.","marker":"Binney & Tremaine 2008"},{"why":"gives the local globular-cluster masses and radii against which the predicted descendants are judged.","marker":"Baumgardt et al. 2023"},{"why":"releases the public version of the N-body code in which all simulations are run.","marker":"Weinberger et al. 2020"},{"why":"provides the method for building the NFW-like halo initial conditions as a matching Hernquist model.","marker":"Springel et al. 2005a"}],"fun_headline_variants":["Sparkler clumps too big to become globular clusters after 9 Gyr","Surviving Sparkler clumps keep 20-50 pc radii, too large for GCs","Simulations show most Sparkler clumps end up too big for globular clusters","Dynamical friction sinks massive clumps, but survivors stay oversized","Sparkler's clumps: only 40% survive, and they're still too large for GCs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that survivors are too large to become globular clusters rests on the measured 30–52 pc effective radii being real physical sizes; if JWST resolution or lensing magnification inflates them, the clumps would be denser, tidal stripping would remove less mass, and the over-massive survivors would persist.","fun_headline_variants_meta":{"raw":{"variants":["Sparkler clumps too big to become globular clusters after 9 Gyr","Surviving Sparkler clumps keep 20-50 pc radii, too large for GCs","Simulations show most Sparkler clumps end up too big for globular clusters","Dynamical friction sinks massive clumps, but survivors stay oversized","Sparkler's clumps: only 40% survive, and they're still too large for GCs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001193,"raw_usage":{"total_tokens":5057,"prompt_tokens":1216,"completion_tokens":3841,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":832,"completion_tokens_details":{"reasoning_tokens":3729}},"tokens_in":832,"tokens_out":3841,"duration_ms":29266,"temperature":1.0,"reasoning_tokens":3729,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T16:14:25.663075+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-estimate each clump's lensing magnification and intrinsic size from a deeper lens model or from resolved stellar kinematics: if the true half-mass radii come out near a few parsecs rather than 30–50 pc, the size-based conclusion fails. An independent check is to search local galaxies of roughly $10^{10}$ solar masses for surviving extended clusters with radii of 20–50 pc and masses above $10^6$ solar masses; their absence would support the paper's prediction.","supporting_citations":[{"cited_title":"G., Desprez , G., et al","cited_arxiv_id":null,"evidence_quote":"discovered the Sparkler galaxy and its system of stellar clumps, defining the observational target."},{"cited_title":"2023, , 520, 2180","cited_arxiv_id":null,"evidence_quote":"supplies the ten clump masses, effective radii, and source-plane distances that seed and calibrate the simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes that the clumps' projected distances resemble the disk/halo globular-cluster distribution, motivating the comparison."},{"cited_title":"2020, , 248, 32","cited_arxiv_id":null,"evidence_quote":"releases the public version of the N-body code in which all simulations are run."}],"review_version":1}