{"id":"1737b81d-5fe8-4fcf-b808-d269ab4da83c","arxiv_id":"2608.12132","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In the COLIBRE simulations, interacting galaxies show a specific star formation rate enhancement of up to about 1.7 at separations of 5-10 kpc, declining to unity by ~200 kpc, and underpredicting SDSS enhancements by about a factor of 2.","lead":"Galaxy interactions in the new COLIBRE simulations boost the rate at which galaxies form stars, by up to about a factor of two at separations near 10 kiloparsecs, fading by about 200 kiloparsecs. The simulated boost is about half as strong as SDSS observations show, and it adds only a few percent to the total cosmic star formation rate.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Stellar mass stripping at close separations may inflate the sSFR enhancement; the paper does not test whether Q(sSFR) reflects increased SFR or reduced M*.","rationale":"The reader identified residual environmental differences as the weakest assumption, but the paper provides direct evidence against this for the fiducial sample: Fig. 2 shows Q(sSFR) converges to unity at large separations for M*>10^10 galaxies, and Fig. 12 demonstrates that sSFRs of interacting and control galaxies converge at early times when pair separations are large. These tests suggest environmental residuals do not dominate the fiducial result. In contrast, the stripping effect is directly measured in Appendix D at exactly the separations where the enhancement peaks, and no test is provided to assess its impact on Q(sSFR). A 20 per cent mass loss at r ≈ 5–10 kpc would inflate sSFR by ≈25 per cent, potentially reducing the headline Q from ≈1.7 to ≈1.4. This is a concrete, unresolved quantitative bias. The paper's supporting evidence (enhanced molecular gas fractions in Fig. 6, and the positive 2.1 per cent contribution to cosmic SFR density using SFR directly) indicates that some genuine SFR enhancement exists, so the concern does not warrant rejection; however, the quantitative amplitude of the central claim and its physical interpretation depend on tests the paper has not performed. The verdict remains CONDITIONAL, pending a direct check of Q(SFR) or a stripping-corrected Q(sSFR). The reader's emphasis on environment is defensible, but the stripping concern is more load-bearing because it is unaddressed and directly tied to the headline amplitude.","tokens_in":35939,"tokens_out":13461,"duration_ms":118468,"concrete_test":"Compute Q(SFR) = <SFR(interacting)>/<SFR(control)> in the same separation bins as Fig. 2 for the L200m6 M*>10^10 sample. If Q(SFR) is substantially lower than Q(sSFR) at r3D < 20 kpc, then the sSFR enhancement is partly due to stellar mass loss rather than increased star formation. As a second check, recompute Q(sSFR) using the maximum stellar mass within the last 500 Myr (M_max*, as defined in Appendix D) for both interacting and control galaxies; a drop of more than 15 per cent in the lowest separation bin would indicate that the reported 1.7 enhancement is inflated by stripping.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that interactions enhance sSFR by up to ≈1.7 at r3D ≈ 5–10 kpc assumes that the measured sSFR increase is driven by elevated star formation rather than by a decrease in the denominator, stellar mass. Appendix D (Fig. D1) shows that interacting galaxies lose up to ≈20 per cent of their stellar mass at the smallest separations (r_sep ≈ 1), due to combined physical and numerical stripping. Since sSFR = SFR/M*, a 20 per cent reduction in M* within the 10 kpc aperture inflates sSFR by ≈25 per cent even if SFR is unchanged. The paper explicitly does not correct for this stripping (Section 4.4, Appendix D) and does not test its impact on Q(sSFR). The time-evolution test in Fig. 12 cannot distinguish stripping from star formation, because both are transient effects of the close passage; the control galaxies are less affected by stripping, so the ratio is biased high. The matching procedure likewise does not address this within-galaxy mass redistribution. If the true enhancement in SFR itself is substantially lower than that in sSFR, the headline factor of 1.7 overstates the effect of interactions on star formation. This concern is directly evidenced at the exact separations where the enhancement peaks, unlike the residual-environment concern, which is shown to converge for the fiducial M*>10^10 sample (Fig. 2).","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the COLIBRE cosmological hydrodynamical simulations at z≈0 to measure the mean specific star formation rate enhancement Q(sSFR) of interacting galaxies relative to mass-, environment-, and redshift-matched isolated controls. It reports Q≈1.7 at separations of ≈5 kpc for galaxies with M*>1e10 M_sun, enhancement remaining significant out to ≈200 kpc, stronger enhancement for lower-mass galaxies and smaller apertures, a resolution-dependent normalization, a comparison with SDSS showing a similar separation dependence but a normalization lower by about a factor of two, and a ≈2.1 per cent contribution of pre-merger interactions to the z≈0 cosmic SFR density.","tokens_in":36226,"tokens_out":6855,"duration_ms":63616,"significance":"If the results hold, they provide a quantitative benchmark for merger-driven star formation in a modern simulation with a multiphase interstellar medium. The SDSS comparison is a genuine prediction, because the COLIBRE model was calibrated to the stellar mass function and size-mass relation rather than to interaction-induced sSFR enhancement. The bootstrap uncertainties and the box-size/resolution convergence tests in Section 4.2 support the statistical claims. The main qualification is the denominator confound from stellar mass stripping, which directly affects the headline amplitude of Q(sSFR), and the need to align the abstract with the fiducial 10 kpc aperture numbers.","major_comments":[{"comment":"The paper measures sSFR within a 10 kpc aperture and interprets Q(sSFR) as the interaction-induced enhancement of star formation, but it does not test whether the enhancement is driven by SFR or by a decrease in the stellar-mass denominator. Equation (2) defines Q(sSFR) as the ratio of mean sSFRs, and Appendix D (Fig. D1) shows that the mean stellar mass ratio M_now/M_max decreases by up to ≈20 per cent at r_sep≈1 due to combined physical and numerical stripping. Section 4.4 declines to apply the Patton et al. (2020) correction and does not present Q(SFR). For a galaxy with 20 per cent mass loss, an observed Q(sSFR)≈1.7 would correspond to Q(SFR)≈1.36 if the SFR were unchanged. The authors should compute the SFR-only ratio Q(SFR), or apply a stripping correction, or at least quantify the maximum effect of stripping on the smallest-separation bin before claiming an sSFR enhancement of the quoted amplitude.","section":"Section 4.4, Appendix D, Eq. (2)"},{"comment":"The abstract states that the average sSFR of interacting galaxies is enhanced by up to a factor of ≈2 for separations of ≈10 kpc, and §3.2.1 says the fiducial 3D sample reaches Q≈2. However, the fiducial result for M*>1e10 M_sun with the standard 10 kpc aperture is Q≈1.7 at r3D≈5 kpc (Fig. 2 and Fig. 8), and Q values near 2 are obtained only for lower stellar masses (Fig. 4) or for apertures of ≤3 kpc (Fig. 8). The abstract and Section 3.2.1 should be harmonized with the fiducial numbers, or explicitly state the mass and aperture conditions under which Q≈2 is reached.","section":"Abstract, §3.1.1, §3.2.1, Figs. 2 and 8"},{"comment":"Equation (A1) as printed, w_xi=|x-x_i|/x_tol, is 0 for a perfect match and 1 at the edge of the tolerance range, which is the opposite of the prose statement that a perfect match has weight 1 and the edge has weight 0. Since the text then says the control with the largest weight is selected, the printed formula would select the worst-matched control. This is either a typographical error (likely missing '1 -' before the ratio) or an inconsistency with the actual matching code. The formula and the selection rule must be corrected and brought into agreement.","section":"Appendix A, Eq. (A1)"}],"minor_comments":[{"comment":"The text says the fiducial 3D sample reaches Q≈2, but Fig. 2 and Section 3.1.1 report Q≈1.7 for the M*>1e10 sample; please specify which stellar mass range and aperture are used in Fig. 9, or change the wording to avoid confusion.","section":"Section 3.2.1, Fig. 9"},{"comment":"There are minor typesetting issues with missing spaces, e.g. 'colibresimulations' in the abstract; a careful proofread would improve readability.","section":"Abstract and text throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope and the central analysis is defensible, but the stripping confound directly affects the headline amplitude and the abstract overstates the fiducial result. The weight-equation inconsistency in Appendix A should also be checked against the actual code, because if the code follows the printed formula the control selection is inverted. With those points addressed, the paper would be suitable for publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper right away. First, it is the first systematic study of interaction-induced sSFR enhancement in the new COLIBRE simulations, and it does a lot of things right. Second, the headline number is shaky: the abstract says “up to a factor of ≈2” but the fiducial 10 kpc aperture gives Q≈1.7 at best, and the paper does not correct for stellar mass stripping that can inflate sSFR by ~25% at the same separations where the enhancement peaks.\n\nThe genuinely new material: COLIBRE includes a multiphase ISM with explicit cold molecular gas, and the paper characterizes how the enhancement depends on separation, stellar mass, mass ratio, aperture, and resolution. The control matching follows the established Patton et al. approach, and the authors are careful to show box-size convergence and resolution trends. The comparison to SDSS is refreshingly honest: they reproduce the separation dependence but admit the normalization is underpredicted by a factor of ~2. The estimate that pre-merger interactions contribute ~2% of the z≈0 cosmic SFR density is a reasonable new number. No parameters are fit to the interaction enhancement; the model was calibrated to other observables, so the SDSS comparison is a real prediction.\n\nThe main soft spot is the stripping issue. Appendix D shows that at the smallest physical separations, interacting galaxies lose up to ~20% of their stellar mass within the aperture, due to combined physical and numerical stripping. Since sSFR = SFR/M*, this alone inflates the enhancement by ~25% even if SFR is unchanged. The paper acknowledges the effect but does not correct for it or test its impact on Q(sSFR). The time-evolution test in Fig. 12 cannot distinguish stripping from a genuine SFR burst, because both are transient and centered on the close passage. This concern lands exactly on the separations where the paper claims its strongest result, so it is not a minor caveat.\n\nThe abstract overstatement is a separate but real issue: Q≈1.7 for the fiducial aperture is not Q≈2, and the higher values come only from smaller apertures or lower-mass bins. The residual environment mismatch in Appendix B is minor for the fiducial M*>10^10 sample, which converges to unity at large separations; the low-mass bin is discussed honestly. The reuse of galaxy pairs across snapshots may underestimate bootstrap errors, and no data products are released, but those are secondary.\n\nWho should read this? Anyone working on galaxy interactions in simulations or comparing predictions to SDSS-style observations. It deserves serious peer review, but the referee should push for a direct decomposition of Q(sSFR) into SFR enhancement versus M* reduction, and for the abstract to match the fiducial result. The qualitative conclusion that interactions enhance star formation will likely survive, but the quantitative claim needs refinement.","headline":"A careful and honest first look at interaction-driven sSFR enhancement in COLIBRE, but the headline factor is inflated by an uncorrected stellar-mass stripping effect and an over-generous abstract.","tokens_in":36833,"tokens_out":1971,"would_cite":true,"duration_ms":19586,"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":"Using the COLIBRE simulations, the paper shows that galaxies with a close companion have mean specific star formation rates enhanced by up to a factor of about 1.7 at 5-10 kpc separations, with a weaker enhancement out to about 200 kpc.","keywords":["galaxy interactions","star formation enhancement","specific star formation rate","hydrodynamical simulations","COLIBRE","matched control samples","SDSS comparison","cosmic star formation rate density"],"falsifier":"Re-run the L200m6 analysis using controls matched on the local density within 0.8 Mpc instead of 2 Mpc, restricted to galaxies with $M_\\ast>10^{10}\\,\\mathrm{M_\\odot}$; if the small-separation enhancement drops well below $Q\\approx1.7$ or the large-separation plateau persists, part of the claimed interaction effect is environmental rather than caused by the companion.","tokens_in":35701,"feed_emoji":"🌌","tokens_out":9500,"duration_ms":74681,"temperature":0.7,"pith_summary":"Using the COLIBRE cosmological hydrodynamical simulations, this paper asks whether galaxies with a close companion form stars faster than otherwise equivalent isolated galaxies. It builds large samples of interacting galaxies and controls matched in stellar mass, redshift, local environment, and isolation, and finds that the mean specific star formation rate is enhanced by up to a factor of about 1.7 at separations of 5-10 kpc, with a smaller but significant enhancement out to roughly 200 kpc. The enhancement is strongest in the central regions, larger for lower-mass galaxies and for more equal-mass pairs, and grows with numerical resolution. Comparison with SDSS observations shows the same separation dependence but a simulation normalisation about a factor of two lower. The paper concludes that resolved pre-merger interactions contribute about 2.1 per cent of the $z\\approx0$ cosmic star formation rate density.","feed_headline":"Close galaxy encounters boost star formation by up to 70 percent","feed_subtitle":"Matched simulation controls show the effect out to 200 kpc and add about 2 percent of the cosmic star formation rate.","key_machinery":"The central object is the sSFR enhancement ratio $Q(\\mathrm{sSFR})=\\langle \\mathrm{sSFR}_{\\rm interacting}\\rangle/\\langle \\mathrm{sSFR}_{\\rm control}\\rangle$ computed in bins of pair separation. The machinery is the COLIBRE simulation model, which resolves a cold, molecular interstellar medium and forms stars in gravitationally unstable gas, together with the Patton et al. (2016) matching algorithm that pairs each interacting galaxy with isolated controls having the same redshift, stellar mass within 0.05 dex, local density within 10 per cent, and isolation within 10 per cent. The ratio isolates the effect of the companion by construction, and the same samples, with projected separations and fibre-like apertures, are used for the SDSS comparison.","core_discovery":"The central claim is that, in the COLIBRE simulations at $z\\approx0$, the mean specific star formation rate of a star-forming galaxy with a companion whose stellar mass is at least one tenth of its own is enhanced relative to a matched isolated control. The enhancement reaches $Q(\\mathrm{sSFR})\\approx1.7$ at a three-dimensional separation of roughly 5 kpc, declines monotonically with separation, and remains statistically significant out to about 200 kpc before converging to unity. The enhancement is stronger for lower-mass galaxies (peaking near $M_\\ast\\approx10^9\\,\\mathrm{M_\\odot}$), for pairs with more equal stellar masses, and when the star formation rate is measured within smaller apertures, reaching about 2.6 within 1 kpc. The COLIBRE prediction has the same radial shape as the SDSS data but is lower in normalisation by about a factor of two, and the paper estimates that the pre-merger excess contributes about 2.1 per cent of the cosmic star formation rate density at $z\\approx0$.","pith_inferences":["I infer that if the resolution trend from m7 to m6 continues to m5, the small-separation normalisation would rise toward the SDSS value, making much of the reported factor-of-two deficit a numerical-resolution effect rather than a physics disagreement.","The residual environment mismatch shown in Appendix B suggests that part of the apparent enhancement in the lowest stellar mass bin is environmental; a direct test would be to adopt sub-Mpc density matching for all mass bins and check whether the large-separation plateau disappears.","Because the paper counts only the pre-merger phase, including post-merger remnants and very low mass-ratio companions would likely push the cosmic contribution above the quoted 2.1 per cent.","The strong aperture dependence predicts that fibre or IFU observations of galaxy centres should find larger interaction-induced sSFR excesses than integrated galaxy-wide measurements."],"forward_implications":["At separations below about 50 kpc, the sSFR enhancement is large and rises steeply, so close pairs dominate the interaction-driven excess at $z\\approx0$.","The enhancement is concentrated in galaxy centres: measuring within 1 kpc apertures roughly doubles the effect compared with 10 kpc apertures.","Lower stellar mass and more equal-mass pairs show stronger enhancement, while the most massive galaxies show almost no mean enhancement.","COLIBRE matches the observed radial dependence of the SDSS enhancement but predicts about half the amplitude, and the gap narrows at higher resolution.","Pre-merger galaxy interactions contribute only about 2.1 per cent of the cosmic SFR density at $z\\approx0$, so they are not the main driver of cosmic star formation at this epoch."],"supporting_citations":[{"why":"Supplies the control-matching algorithm used to pair each interacting galaxy with isolated galaxies matched in stellar mass, local density, isolation, and redshift.","marker":"Patton et al. (2016)"},{"why":"Provides the SDSS observational sample of interacting and control galaxies used for the comparison of Q(sSFR) as a function of projected separation.","marker":"Patton et al. (2013)"},{"why":"Gives the previous simulation measurements in EAGLE, Illustris, and IllustrisTNG and the 200 kpc detection that this work extends and compares against.","marker":"Patton et al. (2020)"},{"why":"Defines the COLIBRE galaxy formation model and its calibration, including the multiphase ISM and resolution setups used here.","marker":"Schaye et al. (2026)"},{"why":"Documents the COLIBRE simulation suite and subgrid calibrations on which the sample construction rests.","marker":"Chaikin et al. (2026a)"},{"why":"Supplies the HBT-HERONS halo finder used to identify interacting subhaloes and their merger histories.","marker":"Forouhar Moreno et al. (2025)"},{"why":"Explains how galaxies seen at separations near 200 kpc can still show enhanced sSFR through earlier pericentric passages, underpinning the interpretation of the large-separation signal.","marker":"Faria et al. (2025)"}],"fun_headline_variants":["Galaxy interactions boost star formation up to 70% in COLIBRE sims","Close galaxy pairs show 70% higher star formation in simulations","Simulations find galaxy interactions boost star formation 70%","Galaxy encounters show 70% star formation boost in COLIBRE","COLIBRE simulations reveal 70% star formation boost from interactions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the matched control galaxies are statistically identical to interacting galaxies in every star-formation-relevant property except the presence of a close companion; Appendix B shows this is only approximately achieved, with residual differences in small-scale environment for low-mass galaxies.","fun_headline_variants_meta":{"raw":{"variants":["Galaxy interactions boost star formation up to 70% in COLIBRE sims","Close galaxy pairs show 70% higher star formation in simulations","Simulations find galaxy interactions boost star formation 70%","Galaxy encounters show 70% star formation boost in COLIBRE","COLIBRE simulations reveal 70% star formation boost from interactions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00036,"raw_usage":{"total_tokens":2029,"prompt_tokens":1110,"completion_tokens":919,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":726,"completion_tokens_details":{"reasoning_tokens":826}},"tokens_in":726,"tokens_out":919,"duration_ms":8629,"temperature":1.0,"reasoning_tokens":826,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:14:41.814000+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the L200m6 analysis using controls matched on the local density within 0.8 Mpc instead of 2 Mpc, restricted to galaxies with $M_\\ast>10^{10}\\,\\mathrm{M_\\odot}$; if the small-separation enhancement drops well below $Q\\approx1.7$ or the large-separation plateau persists, part of the claimed interaction effect is environmental rather than caused by the companion.","supporting_citations":[],"review_version":1}