{"id":"d3c50367-7bfb-4c24-a924-c4de9cd6ba89","arxiv_id":"2507.02051","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The orbital geometry of a galaxy merger, spiral-in versus direct collision, systematically controls the disk, bulge, and hot halo fractions of the remnant galaxy in IllustrisTNG-100.","lead":"Using 531 galaxy mergers from a cosmological simulation, this paper shows that the way two galaxies approach each other, spiraling in gently versus colliding head-on, systematically controls the structure of the merged galaxy. The result helps explain why some galaxies end up with big hot halos and why common structural signs of galaxy mergers are so scattered.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported orbit-structure correlations use post-merger fractions, not merger-induced changes; without controlling for pre-merger structure, the claimed orbital control may be inherited from progenitor properties.","rationale":"The reader's t±1Gyr assumption is worth checking, but it is only one component of the measurement pipeline, and the paper's manual visual exclusion (§2.3, Appendix A) is a direct, if imperfect, response. My concern targets the inferential step that carries the headline: the correlations in Figs. 6 and 8 compare remnant fractions to orbital angles, not the change caused by the merger. A remnant fraction is not a transformation; it is a mixture of the initial state and the merger effect. If progenitor disk fraction correlates with orbital configuration, then high remnant disk fraction in spiral-in mergers could be because such mergers are preferentially experienced by already disk-dominated galaxies, not because the orbit preserves the disk. The paper itself provides evidence that orbital configuration and progenitor disk orientation are coupled (§3.1, θsum 116° vs 81°), making this confound plausible. The missing control is cheap and decisive, so the verdict should remain conditional rather than accepting the causal claim as established. I agree only partially with the reader because the snapshot-timing issue is real, but the most load-bearing gap is the absence of a change-based or residualized analysis.","tokens_in":17347,"tokens_out":6900,"duration_ms":83724,"concrete_test":"Recompute the central correlations from Figs. 6 and 8 using the merger-driven change Δf* = f*,+1Gyr − f*,main@−1Gyr as the dependent variable, and also using residualized remnant fractions. Concretely: (1) compute R(⟨θr,v⟩, Δfdisk), R(⟨θr,v⟩, Δfwarm), R(⟨θr,v⟩, Δfbulge), and R(⟨θr,v⟩, Δfhalo) for the three mass-ratio bins; (2) fit fdisk,+1Gyr as a linear function of fdisk,−1Gyr, merger mass ratio, and remnant stellar mass, then correlate the residual with ⟨θr,v⟩ and ⟨θ′Lmain,Lorbit⟩. If the Δf* or residual correlations drop to near zero, the orbit-structure relation is dominated by pre-merger progenitor properties; if R(⟨θr,v⟩, Δfdisk) remains ≈0.6, the concern is refuted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5's central claim is that orbital configuration controls the remnant's structural fractions, supported by R(⟨θr,v⟩, fdisk,+1Gyr)=0.66, R(⟨θr,v⟩, fwarm,+1Gyr)=0.66, and weaker anti-correlations for bulge and halo (§3.2, §3.3, Figs. 6 and 8). These correlations are computed against the remnant fraction f*,+1Gyr only, not against the merger-induced change Δf* = f*,+1Gyr − f*,main@−1Gyr that the paper itself defines. Because f*,+1Gyr is strongly bounded by the progenitor's f*,−1Gyr, and because §3.1 shows the orbital configuration is correlated with the main-disk orientation (θsum averages 116° for radial orbits vs 81° for spiral-in orbits for major mergers), the reported dependence could simply reflect that galaxies with higher pre-merger disk fractions preferentially have certain orbital configurations. The paper does not report fdisk,−1Gyr versus ⟨θr,v⟩, nor the partial correlation of fdisk,+1Gyr with ⟨θr,v⟩ after controlling for fdisk,−1Gyr, mass ratio, and gas fraction; Fig. 9 plots Δf* against satellite mass fraction but not against orbital angles. Without this control, the causal reading of the headline result is underdetermined.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript analyzes 531 merger pairs from IllustrisTNG-100, restricting to mergers with mass ratio above 1/5 and main progenitors with disk fraction above 0.1. It decomposes galaxies into disk, warm, bulge, and hot inner stellar halo components using circularity and radius cuts, defines three merger orbital angles (θr,v, θLmain,Lorbit, θLsat,Lorbit), and reports correlations between these angles and the structural fractions of the remnant measured 1 Gyr after the merger. The main claims are that spiral-in/tangential mergers produce remnants with higher disk and warm fractions and lower bulge and hot inner halo fractions, while head-on/radial mergers do the opposite; that the hot inner halo fraction increases in nearly all mergers; and that this accumulated behavior explains the z=0 correlation between inner halo fraction and total ex situ stellar mass. The paper also proposes spiral-in mergers as a possible formation channel for red but HI-rich galaxies, based on two simulated counterparts.","tokens_in":17579,"tokens_out":10092,"duration_ms":121381,"significance":"If the causal interpretation is upheld, the paper would establish orbital configuration, not only mass ratio or gas fraction, as a first-order determinant of merger-driven structural transformation, and would connect single-merger orbital properties to the scatter in structural indicators of ex situ mass at z=0. The work is based on a public simulation, uses a clearly defined sample and explicitly defines merger-induced changes Δf* even though the headline correlations are not computed on those changes. The proposed RR-galaxy channel is interesting but rests on two objects and is appropriately framed as a possibility. The main weakness is that the causal reading of the central correlations is underdetermined without controlling for progenitor structure, so the significance is conditional on additional analysis.","major_comments":[{"comment":"The central causal claim—that orbital configuration reshapes remnant structure—is based on correlations between the orbital angle ⟨θr,v⟩ and the post-merger fractions f*+1Gyr (R = 0.66, 0.66, −0.36, −0.30 for major mergers), not on correlations with the merger-induced change Δf* = f*+1Gyr − f*main,-1Gyr that is defined in Section 3.2. This distinction matters because Section 3.1 and Summary item 1 show that the orbital configuration is itself correlated with the pre-merger disk orientation: for major mergers, θsum averages 116° on radial orbits versus 81° on spiral-in orbits. If high-disk-fraction progenitors preferentially end up on spiral-in orbits, the reported f*+1Gyr–angle relations could simply reflect initial conditions rather than the merger itself. Please add (i) plots and Pearson/Spearman correlations of Δfdisk, Δfwarm, Δfbulge, and Δfhalo versus ⟨θr,v⟩; (ii) partial correlations of f*+1Gyr with ⟨θr,v⟩ controlling for f*,-1Gyr, merger mass ratio, and gas fraction; and (iii) fdisk,-1Gyr versus ⟨θr,v⟩ and θsum. Without these, the abstract's 'lead to higher fractions' is not supported by the presented statistics.","section":"Section 3.2, Figs. 6 and 8"},{"comment":"The visual filter that excludes pairs with pre-merger disturbance or post-merger substructure is described only by four representative examples and no count. Because the t+1Gyr snapshot is the operational definition of the settled remnant for all 531 pairs, and because the likelihood of being out of equilibrium could plausibly depend on orbital configuration, this subjective filter can bias all of the reported correlations. Please quantify the filter: report how many pairs were excluded, give the distribution of the three angles for excluded versus included pairs, and test sensitivity to an automated equilibrium criterion such as the substructure mass fraction or the convergence of the component fractions between t+1Gyr and t+2Gyr.","section":"Section 2.3 and Appendix A"},{"comment":"The Pearson correlation coefficients are reported without uncertainties, p-values, or a statement about whether they are computed on individual points or on the binned running medians. Given the modest values for bulge and halo (R = −0.36 and −0.30) and the highly non-Gaussian angle distributions shown in Fig. 6, the paper should provide bootstrap confidence intervals and Spearman rank coefficients so that the relative strength of the correlations, and in particular the claim that 'the correlations are much weaker in minor mergers', is quantitatively grounded.","section":"Section 3.2 and 3.3, Figs. 6 and 8"}],"minor_comments":[{"comment":"The disk component is defined as λz > 0.8 without a radial restriction, while warm, bulge, and halo are confined to rcut < r < rmax or r < rcut. Please clarify whether the four fractions are normalized to total stellar mass and whether they are intended to partition the galaxy; as written, stellar mass outside rmax with λz > 0.8 is counted only in fdisk, so the four fractions need not sum to unity.","section":"Section 2.2"},{"comment":"The abstract quotes 93% for the hot inner stellar halo increase, but Section 3.3 only says 'nearly always increases'; please state where the 93% is obtained or remove the number from the abstract.","section":"Abstract and Section 3.3"},{"comment":"The claimed strong positive correlation between fhalo,z=0 and the ex situ stellar mass fraction is not quantified; please add a correlation coefficient, a fit, or a scatter estimate so that the strength of the relation can be compared with the scatter shown in the panel.","section":"Fig. 9 bottom panels"},{"comment":"The captions contain non-English placeholder text ('给Sample exhibition添加Redshift z' and '* theta_r,v调整为被90度折叠后的') that should be removed before submission.","section":"Figs. 3 and 5"},{"comment":"Equation 5 and the citation to Cai et al. (2013) appear questionable as the source for the isotropic PDF p(θ) = sin θ; please verify the reference or cite a standard source for the isotropic distribution of 3D angles.","section":"Section 3.1"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the missing Δf analysis; if the authors supply it and the results survive, I would support acceptance. The reliance on Zhu et al. (2022a) for the decomposition thresholds is appropriate, and the paper is within the scope of the journal. The RR-galaxy section is speculative but clearly framed, so it should not block publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Best read as a systematic phenomenological study, not a proof of causal mechanism. The authors map three orbital angles to four dynamically-decomposed components across 531 merger pairs in a public simulation. That alone is a useful step beyond disk-survival studies like Zeng et al. and Sotillo-Ramos et al. The main result—spiral-in orbits correlate with higher disk/warm fractions and lower bulge/halo fractions—is consistent with earlier work, and the correlations are moderate but real. I like that they checked the disk definition systematically and that the manual exclusions are documented.\n\nThe soft spot is the one flagged in the stress test: the headline correlations are computed against f_+1Gyr, not against the change Δf = f_+1Gyr − f_main,−1Gyr. The paper defines Δf and even colors points by it, but the running medians and Pearson coefficients use the post-merger value. Since f_+1Gyr carries memory of the progenitor's structure, and since orbit correlates with main-disk orientation (θsum averages 116° for radial vs 81° for spiral-in), some of the correlation could be inherited. A partial correlation controlling for f_disk,−1Gyr and mass ratio would settle this. It's a fixable weakness, not a fatal one—the Δf trends in the top of Fig. 9 are suggestive, but they're not quantified against angle.\n\nMinor concerns: no error bars on correlation coefficients, the visual exclusion of merger pairs is subjective (though documented), and the RR galaxy discussion rests on two simulated analogues, which is explicitly speculative. The inner halo–ex situ correlation is already known from Zhu et al. (2022a), so that part is confirmatory, not new. Self-citation is not the issue here.\n\nOverall, a useful and well-organized paper that deserves referee time. I'd ask for the Δf-based correlations and partial controls before accepting; the central direction of the effect is likely right, but the causal language in Section 5 should be toned down until then.","headline":"A careful systematic study of merger orbits and remnant structure in TNG100; the headline correlation is plausible but the causal claim needs a cleaner pre-merger control.","tokens_in":18150,"tokens_out":2038,"would_cite":false,"duration_ms":24085,"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":"Merger orbital geometry systematically reshapes remnant galaxies, with spiral-in orbits preserving disks and head-on collisions building hot inner halos.","keywords":["galaxy mergers","orbital configuration","galaxy structure decomposition","disk survival","hot inner stellar halo","ex situ stellar mass","IllustrisTNG","red HI-rich galaxies"],"falsifier":"Recompute the component fractions for the same 531 merger pairs at $t_{\\rm merger}+2$ Gyr and $t_{\\rm merger}+3$ Gyr. If the ordering of remnants by orbital configuration changes materially between one and three gigayears after the merger, or if the $t-1$ Gyr progenitors show tidal distortions that grow with the radiality of the orbit, the claimed orbit–structure relation is an artifact of the chosen snapshot times.","tokens_in":17110,"feed_emoji":"🌀","tokens_out":9709,"duration_ms":92377,"temperature":0.7,"pith_summary":"This paper tries to establish that the orbit along which two galaxies merge—not just their mass ratio or gas content—systematically decides what structure the remnant galaxy ends up with. Using 531 merger pairs from the TNG100 cosmological simulation, the authors describe each merger by the relative motion of the pair and the orientations of both galactic disks, then dynamically split every galaxy into disk, warm, bulge, and hot inner stellar halo components. They find that spiral-in orbits align the progenitor disks with the orbital plane and leave remnants rich in disk and warm components, while head-on collisions produce bulge-dominated remnants with large hot inner halos. Almost every merger, regardless of orbit, increases the hot inner stellar halo fraction, which is why that component (unlike the bulge) tracks total accreted stellar mass at $z=0$. A sympathetic reading takes this as evidence that merger geometry is a first-order variable in galaxy evolution, not a secondary correction.","feed_headline":"Merger orbit, not mass ratio, sets remnant structure","feed_subtitle":"Spiral-in collisions preserve disks and warm stars; head-on crashes build bulges and hot inner halos.","key_machinery":"There are two pieces of machinery. The first is a dynamic decomposition of each galaxy into four components using the circularity $\\lambda_z = J_z/J_{\\max}(E)$ and radius: disk ($\\lambda_z > 0.8$), warm component ($0.8 > \\lambda_z > 0.5$ outside the bulge radius), bulge ($\\lambda_z < 0.8$, $r < 3.5$ kpc), and hot inner stellar halo ($\\lambda_z < 0.5$ outside the bulge radius). The second is a three-angle description of the merger: $\\langle \\theta_{r,v}\\rangle$ measures how radial the infall is (zero for head-on, ninety degrees for spiral-in), while $\\langle \\theta'_{L_{\\rm main},L_{\\rm orbit}}\\rangle$ and $\\langle \\theta'_{L_{\\rm sat},L_{\\rm orbit}}\\rangle$ give the orientations of the two galactic spins relative to the orbital plane, folded to 0–90 degrees and summed into $\\theta_{\\rm sum}$. The orbital angle $\\langle \\theta_{r,v}\\rangle$ does the main carrying: it correlates most strongly with the remnant's disk and warm fractions, and it also correlates with $\\theta_{\\rm sum}$, showing that orbital geometry and galaxy orientation are tied together in the merger population.","core_discovery":"The paper's central claim is that the orbital configuration of a merger is a first-order determinant of the remnant's structure. For mergers on spiral-in orbits the disk planes of both galaxies tend to lie nearly parallel to the orbital plane, and the remnant retains higher fractions of disk and warm components and lower fractions of bulge and hot inner stellar halo; for mergers on head-on collision orbits the disk planes tend to lie perpendicular to the orbital plane, producing the opposite outcome. The correlation between the orbital angle $\\langle \\theta_{r,v}\\rangle$ and remnant disk fraction reaches $R=0.66$ for major mergers, comparable in strength to mass-ratio effects. A second load-bearing result is that mergers almost always grow the hot inner stellar halo, increasing its fraction in 93% of cases, so after several mergers a galaxy's inner-halo fraction becomes a faithful record of its total accreted stellar mass. This is offered as an explanation for the observed correlation between inner-halo luminosity and ex situ stellar mass at $z=0$, and as a caution that disk or bulge mass fractions cannot be read as universal merger counters.","pith_inferences":["Beyond the paper: if orbit is this influential, semi-analytic and zoom-in merger models that track only mass ratio and gas fraction will mispredict remnant structure; adding the three angles as inputs would sharpen morphology predictions without new physics.","Beyond the paper: the angle–structure relations could be tested observationally by classifying mergers as prograde versus retrograde from stellar kinematics and comparing the resulting halo and disk properties in the local universe.","Beyond the paper: the 93% hot-halo increase suggests that measuring the hot inner stellar halo of a galaxy could serve as a practical single-number proxy for accumulated merger mass, even when the full merger tree is unavailable.","Beyond the paper: a direct check would be to rerun a subset of the same mergers at higher time resolution or in a different simulation code; if the orbit–structure ordering survives, the result is robust to numerical details."],"forward_implications":["Remnant disk survival in the simulation is predictable from the orbital angle: tangential, spiral-in accretion preserves disks, radial head-on accretion destroys them, with correlation $R=0.66$ for major mergers.","Because the hot inner stellar halo fraction rises after nearly every merger, it accumulates accreted mass across a galaxy's whole history, making it a more dependable indicator of ex situ stellar mass than bulge fraction.","The $z=0$ correlation between inner-halo fraction and ex situ stellar mass emerges from the averaging of many individual mergers, each of which scatters widely around the relation.","Spiral-in mergers can produce red but HI-rich galaxies by depositing HI at large radius with high angular momentum while the stellar body remains quenched.","Disk and bulge fractions should not be used as simple merger counters, since a merger can raise or lower them depending on orbit."],"supporting_citations":[{"why":"supplies the four-component dynamical decomposition used for every galaxy and the known correlation between inner-halo fraction and ex situ stellar mass that this paper explains.","marker":"Zhu et al. (2022a)"},{"why":"shows disk survival depends on the collision angle between velocity and position vectors, the precedent this paper extends to all four components.","marker":"Zeng et al. (2021)"},{"why":"shows that combining collision angle and mass ratio predicts disk survival, supporting the use of angle parameters.","marker":"Sotillo-Ramos et al. (2022)"},{"why":"provides the baryonic subhalo merger trees used to identify the 531 merger pairs and their progenitors.","marker":"Rodriguez-Gomez et al. (2015)"},{"why":"presents the TNG simulation suite that produced TNG100, the source of all galaxy properties in the sample.","marker":"Springel et al. (2018)"},{"why":"reports the red but HI-rich galaxies that the paper seeks to explain via spiral-in mergers.","marker":"Li et al. (2024)"}],"fun_headline_variants":["Orbital geometry, not mass ratio, decides remnant structure","Spiral-in mergers preserve disks; head-on collisions build bulges","Merger orbits set disk versus bulge fractions","Hot inner halo grows in 93% of mergers, tracks accreted mass","Orbital alignment decides remnant's disk and bulge mix"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that one gigayear before a merger the two galaxies are still unperturbed and that one gigayear after it the remnant has already settled into equilibrium; if either timescale is wrong, the measured structural changes would not reflect the merger's permanent outcome.","fun_headline_variants_meta":{"raw":{"variants":["Orbital geometry, not mass ratio, decides remnant structure","Spiral-in mergers preserve disks; head-on collisions build bulges","Merger orbits set disk versus bulge fractions","Hot inner halo grows in 93% of mergers, tracks accreted mass","Orbital alignment decides remnant's disk and bulge mix"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000557,"raw_usage":{"total_tokens":2701,"prompt_tokens":1044,"completion_tokens":1657,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":660,"completion_tokens_details":{"reasoning_tokens":1572}},"tokens_in":660,"tokens_out":1657,"duration_ms":15162,"temperature":1.0,"reasoning_tokens":1572,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:39:30.856323+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the component fractions for the same 531 merger pairs at $t_{\\rm merger}+2$ Gyr and $t_{\\rm merger}+3$ Gyr. If the ordering of remnants by orbital configuration changes materially between one and three gigayears after the merger, or if the $t-1$ Gyr progenitors show tidal distortions that grow with the radiality of the orbit, the claimed orbit–structure relation is an artifact of the chosen snapshot times.","supporting_citations":[{"cited_title":"2022, Monthly Notices of the Royal Astronomical Society, 516, 5404","cited_arxiv_id":null,"evidence_quote":"shows that combining collision angle and mass ratio predicts disk survival, supporting the use of angle parameters."},{"cited_title":"2015, Monthly Notices of the Royal Astronomical Society, 449, 49","cited_arxiv_id":null,"evidence_quote":"provides the baryonic subhalo merger trees used to identify the 531 merger pairs and their progenitors."},{"cited_title":"2018, Monthly Notices of the Royal Astronomical Society, 475, 676","cited_arxiv_id":null,"evidence_quote":"presents the TNG simulation suite that produced TNG100, the source of all galaxy properties in the sample."}],"review_version":1}