{"id":"b3820aa4-60f4-4b84-9b8e-3a61d2e68ede","arxiv_id":"2501.07097","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Protoclump regions in VELA simulations show over-compressive tidal fields and 2-10 times excess stream material, both correlating with an excess of converging turbulence.","lead":"In simulated high-redshift disk galaxies, the cloudy regions that later become giant star-forming clumps sit in unusually squeezing gravity fields and along incoming gas streams. This suggests clump birth may be triggered from outside, not only by the disk's own gravitational instability.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Tidal tensor includes the protoclump's own density: ftides excess may be self-induced, not external compressive tides.","rationale":"Agree with the reader that the weakest assumption is the interpretation of the local Hessian as an external environmental tracer. This is the single most load-bearing issue because the paper's case for a new external mode of instability rests on the tidal mechanism; the stream mechanism is acknowledged to be crude but is not called into question by this concern. The Q>>1 stability argument does not remove the local density term from ∇²φ. A concrete density-subtraction test would settle whether the tidal signal is external or self-induced. The paper is otherwise careful and includes an isolated-simulation comparison, which is genuine supporting evidence, but that comparison does not control for the density contrast of VELA protoclumps. Since the paper already carries a CONDITIONAL verdict due to causal interpretation, this concern sharpens the required condition but does not change the verdict.","tokens_in":26701,"tokens_out":7798,"duration_ms":75857,"concrete_test":"Recompute ftides for protoclumps and random patches using the Hessian of the gravitational potential with the contribution of all mass inside each protoclump sphere (R<0.5 kpc) removed (e.g., by solving the Poisson equation with the density inside the sphere set to zero, or by direct summation of external particles). If the protoclump excess in ftides and the fraction of λ3>0 persist, the compressive tides are external; if they vanish or strongly decrease, the tidal signal is dominated by the protoclump's own density enhancement and the causal interpretation in Section 5 is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.4 defines the tidal tensor as the Hessian of the full gravitational potential, T_ij = ∂²φ/∂x_i∂x_j, and Section 3.2.2 interprets positive ftides (eq. 14) as compressive tides from the cosmological environment. Because ∇²φ = 4πGρ, the trace (and the individual eigenvalues) of the Hessian inside any region contain the local density of that region itself. Protoclumps are, by construction, density enhancements (the clump finder selects δ > 10 regions; Section 2.2). For a roughly spherical overdensity, the self-contribution to the Hessian is positive in all three directions (e.g., a uniform sphere gives λ1=λ2=λ3=4πGρ/3 and ftides=2). Therefore ftides > 0 and even fully compressive tides (λ3>0) are expected for any local density peak, independent of the external environment. The argument in Section 3.2.2 that Q>>1 implies the protoclump is 'not self-gravitating' does not address this: Q is a stability criterion involving velocity dispersion and surface density, not a measure of the local density contribution to the tidal tensor. A high-σ region can be a mild overdensity and not collapse while still producing positive tidal eigenvalues. Thus the observed ftides enhancement and the 25% fully compressive fraction may be a consequence of the protoclump's own density enhancement (or of the collapse process itself) rather than evidence for external compressive tides. The comparison of random patches, which are not density peaks, to protoclumps, which are, does not control for this. The isolated-simulation comparison in Section 4.1 mitigates but does not eliminate the concern, since protoclumps in VELA could be denser relative to their surroundings.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies protoclump regions in eight VELA3 cosmological zoom-in simulations, testing two candidate external drivers of the excess compressive turbulence found earlier in these protoclumps: compressive gravitational tides from the cosmological environment and direct driving by inflowing gas streams. The authors define three diagnostics: fconv, the local fraction of kinetic energy in converging modes of turbulence; ftides, a dimensionless measure of how compressive the gravitational tidal tensor is; and fstr, the local excess of stream material relative to the azimuthal average at the same galactocentric radius. They compare protoclumps to matched random disk patches and report that protoclumps have higher fconv, preferentially positive ftides (median ~0.33, with ~25% fully compressive), and enhanced fstr (median ~2.5), with Spearman correlations of 0.46 (ftides vs fconv) and 0.30 (fstr vs fconv). The paper concludes that compressive tides and stream-disk interactions can drive the compressive turbulence that initiates clump formation in disks where Toomre Q is high.","tokens_in":83,"tokens_out":7084,"duration_ms":132145,"significance":"If the causal interpretation were established, the paper would provide a new, plausible mechanism for giant clump formation in high-z disks that is complementary to classical Toomre instability, and it would connect clump formation to cosmological processes (tides and cold streams). The work extends the earlier single-galaxy analysis of Mandelker et al. (2025) to eight galaxies and introduces a quantitative, local definition of tidal compressiveness (ftides) that is clearly explained. The authors are transparent about many caveats, including the absence of tracer particles in the stream analysis and the correlational nature of the analysis. However, the central evidence for the tidal mechanism is weakened by a selection effect that has not been controlled for, as detailed below; the stream analysis is more robust but still crude. The paper is clearly written and would be of interest to the community if the tidal result can be made self-contamination-free.","major_comments":[{"comment":"The tidal tensor is defined as the Hessian of the full gravitational potential, so through Poisson's equation it necessarily contains the local density enhancement of the protoclump itself. Since protoclumps are defined as δ > 10 density peaks, their self-gravity contributes positive eigenvalues in all three directions (a uniform sphere gives λ1=λ2=λ3=4πGρ/3 and ftides=2). Therefore the median ftides > 0 and the 25% fully compressive fraction may be a direct consequence of the density selection used to identify protoclumps, not evidence of external compressive tides. The argument in §3.2.2 that Q >> 1 implies the protoclump is not self-gravitating does not address this, because Toomre Q is a stability criterion involving velocity dispersion and surface density, not a measure of the local density contribution to the Hessian. The random-patch comparison is also not a valid control, since random patches are not selected to be density peaks. The isolated-simulation comparison in Fig. 8 still shows a positive median ftides ~ 0.2 in protoclumps, which is consistent with a self-gravity floor, and the absence of λ3 > 0 there may reflect different density contrasts or simulation setups rather than the absence of external tides. The authors should recompute the tidal tensor after removing or smoothing the protoclump's own mass (e.g., using a potential computed from the density field smoothed on scales larger than RPC) or quantify and subtract the self term.","section":"§2.4, §3.2.2, Fig. 5"},{"comment":"The causal language in the abstract and conclusions (\"can thus serve as the drivers of excessive compressive turbulence\") goes beyond what the correlations establish. A compressive converging flow (which is exactly what a high fconv means) compresses gas and raises the local density, which in turn raises the self-gravity contribution to ftides; thus the ftides-fconv correlation may be partly a physical consequence of the same converging motion rather than evidence that tides drive that motion. The paper states in §3.3 that causality is not straightforward to establish, but the concluding sections present the correlation as support for a driver role. Please either soften the causal claims or add a time-lagged test, e.g., measuring ftides at an earlier snapshot before the converging flow develops, or comparing regions with similar density but different ftides.","section":"§3.3, Fig. 7, Conclusions"},{"comment":"The stream indicator fstr relies on approximating fluid-element trajectories with a snapshotted velocity field held constant over a disk dynamical time, and it is a mass-based proxy without tracer particles. The paper acknowledges these limitations, but the quantitative claims (70% of protoclumps are stream-interaction sites, fstr = 2–10) are sensitive to the choices of the 10% disk-radius threshold, the backtracking time, and the angular bin size, none of which are varied here. Since protoclumps are dense and may be associated with slow, non-circular flows, the streamline method could systematically misclassify a fraction of the dense gas as 'stream' material. Please report a sensitivity test over these parameters, and ideally compare with a tracer-based identification if any such testbed is available.","section":"§2.5, §3.2.3"}],"minor_comments":[{"comment":"The right-panel caption and x-axis label contain an apparent typo ('log (20 3)' instead of 'log(20λ3)'), and the explanation of how negative λ3 values are encoded on the logarithmic axis is hard to follow; consider plotting λ3 directly or using a two-panel presentation for positive and negative values.","section":"Fig. 5 caption"},{"comment":"The phrase 'maximal resolution' should be 'maximum resolution' for standard English usage.","section":"§2.1"},{"comment":"The text reports a median ftides of 0.32 for protoclumps and −0.26 for random patches, while the figure caption quotes 0.33 and −0.27; please make the numbers consistent.","section":"§3.2.2 and Fig. 5"},{"comment":"The p-values for the Spearman correlations appear only in the Fig. 7 caption and are missing from the body text; include them in the text where the correlation coefficients are reported.","section":"§3.3"},{"comment":"The term 'disk dynamical time' is used but not explicitly defined; give the formula or reference used for this timescale.","section":"§2.5"}],"recommendation":"major_revision","confidential_remarks":"The paper is generally well written and transparent, and the authors clearly distinguish their results from a definitive causal test. The main obstacle is the self-gravity contamination of the tidal tensor measurement, which is a load-bearing issue for the paper's title and central claim. This is fixable by recomputing the tidal diagnostics with the protoclump's own mass removed or smoothed out, and such a revision is within the scope of a major revision. The stream analysis is crude but the authors already acknowledge its limitations; adding robustness checks would strengthen it. I would not recommend rejection because the paper's methodological approach is sound in principle and the isolated-simulation comparison shows the authors are aware of the need for controls."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper extends M25 from one galaxy to eight and shows a clear statistical association between protoclump locations and both compressive tides (ftides > 0, ~25% fully compressive) and stream-disk interaction (fstr 2-10 times the average). That is a solid new result, and the ftides definition is a real improvement over using lambda_3 alone. The correlations (Spearman 0.46 and 0.30) are honestly reported, and the paper is careful in Section 3.3 to say that causal conclusions are not straightforward.\n\nThe main soft spot is the stress-test point: the tidal tensor is the Hessian of the full gravitational potential, which necessarily includes the protoclump's own density. For any roughly spherical overdensity, the self-contribution is positive in all three directions, so ftides > 0 and even lambda_3 > 0 are expected from the protoclump itself. The paper's appeal to Q >> 1 does not answer this. Q is a stability parameter involving surface density and velocity dispersion; it is not a statement about the local density term in del^2 phi = 4 pi G rho. The isolated-simulation comparison mitigates the concern but does not eliminate it, because VELA protoclumps may simply be denser relative to their surroundings, making the self-contribution larger.\n\nThe stream indicator is crude (no tracer particles), which the paper openly admits, and there is no released code or data. These are minor in comparison to the tidal self-contamination issue, but they matter for reproducibility.\n\nWho this is for: galaxy formation people working on clump formation and violent disk instability. The correlational evidence is strong; the causal case needs more work. I would send this to peer review, with a referee instructed to ask for either a decomposition of the tidal tensor into self and external components (e.g., by recomputing the Hessian after removing the protoclump's own mass) or a controlled experiment that isolates each mechanism. With that, the paper's conclusions would be much more convincing.","headline":"Solid correlational extension to eight galaxies, but the tidal tensor's self-gravity contamination undercuts the causal claim; worth reviewing with a demand to address it.","tokens_in":27590,"tokens_out":3590,"would_cite":true,"duration_ms":33735,"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":"Giant clumps in high-redshift disks can be born from compressive tides and stream-disk collisions, not only from classic Toomre instability.","keywords":["high-redshift galaxies","giant clumps","compressive turbulence","tidal tensor","cold streams","disk instability","Toomre Q","cosmological simulations"],"falsifier":"A simulation with tracer particles that follows protoclumps backward in time could settle the order of events: if the excess converging turbulence and compressive tides appear only after the protoclump's own density enhancement starts to grow, rather than before, the proposed environmental drivers are consequences, not causes. Alternatively, recomputing the tidal Hessian after masking out the protoclump's own mass, and finding that most protoclump regions then have the smallest eigenvalue λ3 below zero, would falsify the claim that external compressive tides precede clump formation.","tokens_in":1864,"feed_emoji":"🌌","tokens_out":2242,"duration_ms":92368,"temperature":0.7,"pith_summary":"High-redshift galactic disks form giant star-forming clumps, and the standard explanation is Toomre instability: the disk fragments where its stability parameter Q drops below unity. This paper argues that in realistic cosmological settings clumps can instead form in regions where Q is well above unity, because the surrounding environment can do the squeezing. Using zoom-in cosmological simulations, it shows that protoclump regions systematically sit in compressive tidal fields, with about 25% of them fully compressive, and at sites where cold streams hit the disk carrying 2-10 times the average stream mass. These environmental drivers correlate with the excess of converging turbulence measured in protoclumps, so the paper proposes them as the origin of the compressive modes that let collapse start before self-gravity takes over. If right, this replaces the linear Toomre picture with a non-linear, environmentally driven mode of violent disk instability.","feed_headline":"Cosmic streams and tides, not disk instability, spawn clumps","feed_subtitle":"Protoclumps sit in compressive tidal fields and stream impact zones where Toomre Q exceeds unity.","key_machinery":"The analysis rests on three local diagnostics computed on a 0.2 kpc grid. fconv is the fraction of turbulent kinetic energy in converging modes, defined as the negative-divergence part of the velocity field divided by the full divergence and curl contributions, specifically |∇·v|²_neg over |∇·v|² plus |∇×v|². ftides is defined as (λ2+λ3)/λ1 using the eigenvalues of the tidal tensor T_ij = ∂²φ/∂r_i∂r_j, where positive eigenvalues mean compression, so positive ftides indicates substantially compressive tides and λ3 > 0 indicates fully compressive tides. fstr is the mass of stream material in a protoclump's angular bin divided by the mean stream mass in the surrounding annulus, with stream material identified by backtracking gas cells along streamlines over a dynamical time. Protoclumps are defined as the 0.5 kpc regions from which tracked clumps collapse, and each protoclump is compared against a random patch at the same galactocentric radius.","core_discovery":"Protoclump regions in the VELA cosmological simulations are not random disk patches. Almost all of them have a positive tidal compression parameter ftides, averaging about 0.32, meaning the local tidal field is substantially compressive along at least two directions, and in about 25% of protoclumps the tidal field is fully compressive, with all three eigenvalues of the tidal tensor positive. No random patch shows fully compressive tides. About 70% of protoclumps reside in stream-disk interaction sites, with stream mass fractions 2-10 times the angular average at the same galactocentric radius, while random patches cluster near a fraction of about 0.8. The fraction of turbulent kinetic energy in converging modes is correspondingly high in protoclumps, with a median fconv near 0.5 versus about 0.21 in random patches, and it rises with both ftides and fstr, with Spearman correlation coefficients of about 0.46 and 0.3 respectively. The paper concludes that compressive tides and inflowing streams can drive the excess compressive turbulence that initiates clump formation, constituting a new non-linear mode of violent disk instability in high-redshift galaxies.","pith_inferences":["If the causal ordering holds, clump formation should be predictable from maps of the tidal tensor and stream geometry alone, before any density threshold is crossed; this could be tested in simulations that mask the protoclump's own mass when computing the Hessian.","Tracking protoclump trajectories with tracer particles would distinguish whether compressive tides or stream impacts lead the process, since the current diagnostics are measured at a single formation snapshot.","Observationally, giant clumps at high redshift should preferentially lie near the projected intersections of cold inflows with the disk, and their internal velocity fields should show an excess of converging relative to solenoidal power.","Because compressive driving raises star formation efficiency, the same environmental drivers may boost star formation inside protoclumps even before collapse, linking this formation channel to the measured clump contribution to total star formation."],"forward_implications":["Clump formation in high-redshift disks can proceed where the Toomre Q parameter is much larger than unity, so linear Toomre stability is not a sufficient criterion in cosmological disks.","The contrast between cosmological and isolated simulations is explained: external tides and streams are present only in the cosmological case, matching the observed excess of compressive turbulence in cosmological protoclumps.","The positive correlations of converging turbulence with both ftides and fstr identify two concrete environmental drivers that a future theory of disk fragmentation must include.","Protoclumps can be recognized by compressive tidal fields and stream-impact sites rather than by low Q alone, which gives simulations and observations a new way to find clump formation sites.","A complementary non-linear theory of violent disk instability, balancing converging modes against solenoidal and shear modes, is needed in place of the Toomre-based picture."],"supporting_citations":[{"why":"Showed that protoclumps in the VELA simulations form where Q exceeds unity, the anomaly this paper sets out to explain.","marker":"Inoue et al. (2016)"},{"why":"Found the excess of compressive turbulence in cosmological protoclumps and its absence in isolated disks, motivating the search for external drivers.","marker":"Mandelker et al. (2025)"},{"why":"Established that fully compressive tides arise in cosmological environments and introduced eigenvalue-based tidal quantification.","marker":"Renaud et al. (2009)"},{"why":"Showed that mergers generate compressive tides and compressive turbulence, the mechanism this paper generalizes beyond mergers.","marker":"Renaud et al. (2014)"},{"why":"Provided the streamline-backtracking method used to identify stream material and define fstr.","marker":"Dutta Chowdhury et al. (2024)"},{"why":"Derived the compressive-tide condition in smooth density profiles that underlies the interpretation of ftides.","marker":"Dekel et al. (2003)"},{"why":"Argued statistically that highly supersonic disks can be unstable even at high Q, supporting collapse without linear Toomre instability.","marker":"Hopkins & Christiansen (2013)"},{"why":"Defined the VELA clump finder, clump tracking, and clump properties that anchor the protoclump definition.","marker":"Mandelker et al. (2017)"},{"why":"Established the compressive versus solenoidal decomposition of turbulence and the density-enhancing effect of compressive driving used to interpret fconv.","marker":"Federrath et al. (2010)"}],"fun_headline_variants":["Cosmic tides and streams, not disk instability, seed clumps","Protoclumps form where Toomre is stable: external drivers at work","Compressive tides and stream impacts trigger clump formation in high-z disks","New mode of disk instability: cosmic streams squeeze clumps into existence","Tides and streams, not Toomre, decide where high-z clumps appear"],"cache_read_input_tokens":29568,"weakest_assumption_plain":"The load-bearing assumption is that the measured squeeze-and-stretch forces around a protoclump come from the galaxy and its surroundings, not from the protoclump's own mass or a nearby disk feature; if local self-gravity dominates the signal, the compressive tides would be a result of collapse rather than its cause.","fun_headline_variants_meta":{"raw":{"variants":["Cosmic tides and streams, not disk instability, seed clumps","Protoclumps form where Toomre is stable: external drivers at work","Compressive tides and stream impacts trigger clump formation in high-z disks","New mode of disk instability: cosmic streams squeeze clumps into existence","Tides and streams, not Toomre, decide where high-z clumps appear"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00055,"raw_usage":{"total_tokens":2711,"prompt_tokens":1116,"completion_tokens":1595,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":732,"completion_tokens_details":{"reasoning_tokens":1497}},"tokens_in":732,"tokens_out":1595,"duration_ms":11417,"temperature":1.0,"reasoning_tokens":1497,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:49:08.825631+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A simulation with tracer particles that follows protoclumps backward in time could settle the order of events: if the excess converging turbulence and compressive tides appear only after the protoclump's own density enhancement starts to grow, rather than before, the proposed environmental drivers are consequences, not causes. Alternatively, recomputing the tidal Hessian after masking out the protoclump's own mass, and finding that most protoclump regions then have the smallest eigenvalue λ3 below zero, would falsify the claim that external compressive tides precede clump formation.","supporting_citations":[{"cited_title":"2025, , 538, L9","cited_arxiv_id":null,"evidence_quote":"Found the excess of compressive turbulence in cosmological protoclumps and its absence in isolated disks, motivating the search for external drivers."},{"cited_title":"M., Naab , T., & Theis , C","cited_arxiv_id":null,"evidence_quote":"Established that fully compressive tides arise in cosmological environments and introduced eigenvalue-based tidal quantification."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Showed that mergers generate compressive tides and compressive turbulence, the mechanism this paper generalizes beyond mergers."},{"cited_title":"2017, , 464, 635","cited_arxiv_id":null,"evidence_quote":"Defined the VELA clump finder, clump tracking, and clump properties that anchor the protoclump definition."}],"review_version":1}