{"id":"8d73c05b-2ef7-4b9e-94ee-f2f6a98bcc45","arxiv_id":"2507.09936","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In post-processed RMHD simulations, the observationally reconstructed luminosity-to-mass ratio of star-forming clumps follows L/M ∝ SFE^1.20, independent of clump mass and initial conditions within the tested range.","lead":"This paper simulates collapsing massive star-forming clumps, converts the results into fake telescope images, and measures the ratio of luminosity to mass, L/M, exactly as observers do. It finds that L/M tracks the star formation efficiency of the clump in a mass-independent power law, giving observers a way to read the evolutionary stage of a clump from far-infrared data alone.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The calibration's zero-point (Eq. 10 intercept) is set by f_acc=0.1, an unconstrained sub-grid accretion-luminosity efficiency, so absolute L/M-to-SFE translation is not robust even though slope may be.","rationale":"The slope and the mass-independence of Eq. 10 are plausible and supported by the internal comparison across the 732 synthetic observations, but the advertised use of Eq. 10 as a direct translator from an observed L/M to an SFE requires a trustworthy absolute zero-point. That zero-point is set by f_acc=0.1. The paper gives only a one-line justification ('in line with Ahmad et al. 2024') for this value and does not propagate its uncertainty. Because the entire calibration pipeline is forward-modeled from sink luminosities, this is not a philosophical objection about missing physics but a quantitative check that can be run with the existing suite. The check is more specific than the reader's ISRF/outflows/HII concern, though both point to the same practical issue: Eq. 10's error bars do not represent its true systematic uncertainty. I would keep the CONDITIONAL verdict: if the f_acc sensitivity is large, the paper still provides a valuable methodology and a relative slope, but its headline numerical calibration should be revised to include f_acc as a systematic parameter; if the sensitivity is small, no change is needed.","tokens_in":19715,"tokens_out":11735,"duration_ms":155407,"concrete_test":"Rerun the post-processing pipeline (Radmc-3d thermal Monte Carlo with the sink luminosities recomputed from Eq. 2 and then Hyper extraction) for the full 732-snapshot suite with f_acc = 0.03, 0.3, and 1.0, keeping all other inputs identical, and refit Eq. 10 for each case. If the intercept changes by more than ~0.15 dex, or if the SFE implied at L/M=10 changes by more than a factor of 2, the absolute calibration must be published with f_acc as a dominant systematic; if both stay within the quoted uncertainties, the concern is retired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper presents Eq. 10 as a quantitative calibration, but its zero-point is fixed by one uncalibrated sub-grid parameter. Eq. 2 sets L_acc = f_acc G M_sink \\dot M_sink / R_star with f_acc described as an 'unknown efficiency factor' and adopted as 0.1. Every synthetic SED, and hence every L/M value, depends on this choice. In the early-to-mid SFE regime, the clump luminosity is dominated by accretion luminosity, so log(L/M) at fixed SFE shifts by approximately log10(f_acc/0.1) if f_acc is wrong; the relative contribution of internal stellar luminosity may also modify the slope. The quoted fit uncertainties (±0.03 in intercept, ±0.02 in slope) are purely statistical and do not include this. A factor-of-2 change in f_acc moves the intercept by ~0.3 dex, and a plausible factor-of-5 change (f_acc=0.5) changes the SFE inferred at L/M=10 from ~1.3e-2 to ~3e-3, directly altering the headline conclusion that the majority of star formation happens after L/M>10. Unlike the missing ISRF/outflows/HII, which Sec. 4 explicitly acknowledges as future work, the f_acc sensitivity is not quantified or caveated anywhere, so the tight error bars overstate the reliability of the absolute calibration.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents the second paper of the Rosetta Stone project. It post-processes 24 radiative magnetohydrodynamic simulations of isolated massive clump collapse (two masses, two Mach numbers, three magnetizations, two turbulent seeds) with Radmc-3d Monte Carlo radiative transfer and generates synthetic Hi-GAL/MIPSGAL-like images from 24 to 500 micron plus 1.3 mm. Source extraction with Hyper and graybody SED fitting are used to reconstruct the observationally defined L/M estimator. The central result is Eq. (10), log(Lbol/M) = 1.20^{+0.02}_{-0.02} log(SFE) + 3.28^{+0.03}_{-0.03}, which the authors claim is independent of clump mass and initial conditions, together with a mass-dependent L/M-age relation. The paper concludes that L/M can be directly translated into SFE and that the majority of star formation activity occurs after clumps enter the L/M > 10 phase.","tokens_in":20032,"tokens_out":5913,"duration_ms":64793,"significance":"If Eq. (10) is robust, the paper provides a valuable semi-empirical calibration connecting a widely used observational evolutionary indicator to a simulation-derived quantity, and the end-to-end pipeline from RMHD simulation to synthetic survey images is a useful methodological template for the community. The relation is not circular: SFE is defined from sink masses and the initial gas mass in the simulation, while L/M is reconstructed from synthetic fluxes via an observer-style SED fit, and the paper explicitly shows deviations from a 1:1 relation in Fig. 6. The main weakness is that the absolute zero-point and, to a lesser degree, the slope rest on unquantified systematics (the accretion-luminosity efficiency f_acc, the dust opacity normalization, and the SED-based mass errors) that are not reflected in the quoted fit uncertainties. The paper deserves publication after a systematic error analysis and a more careful statement of the applicability of the calibration.","major_comments":[{"comment":"The zero-point of the headline calibration is set by the sub-grid accretion-luminosity efficiency f_acc = 0.1, which is introduced as an 'unknown efficiency factor' in Eq. (2) and is never varied or propagated into the final relation. In the early-to-mid SFE regime the accretion luminosity dominates the sink emission, so log(Lbol/M) at fixed SFE shifts by approximately log10(f_acc/0.1). A factor-of-2 change in f_acc moves the intercept of Eq. (10) by roughly 0.3 dex; a factor-of-5 change (f_acc = 0.5) changes the SFE inferred at L/M = 10 from about 1.3e-2 to about 3e-3, directly affecting the conclusion that the majority of star formation occurs after L/M > 10. The quoted +/-0.03 intercept uncertainty reflects only the linmix fit scatter. Please add a sensitivity analysis or an explicit systematic-error term to Eq. (10) and to the abstract's quantitative claims.","section":"Sec. 2.2.1, Eq. (2); Sec. 3.2, Eq. (10)"},{"comment":"The clump masses entering L/M carry 20-40% systematic uncertainties (Sec. 3.2) and, at some stages, larger errors; Appendix B reports about 60% uncertainty at the earliest usable snapshot, improving to about 20% overshoot at late times for the fiducial model. Figure 6 shows a systematic offset between the observed and simulation L/M that the text attributes to the mass estimation method, and this offset is also cited as the reason why the slope of Eq. (10) differs from the steeper slope found in Paper I. In addition, the opacity normalization in Eq. (6), kappa_ref = 0.2061 cm2/g, is a factor of about 2 higher than the 0.1 cm2/g used in Elia et al. (2017); if the latter mass scale is used, all reconstructed masses double and L/M halves. None of these systematics appears in the error budget of Eq. (10). Please provide a systematic error budget for the slope and intercept, or state explicitly that Eq. (10) is calibrated only on the authors' adopted mass and opacity scale.","section":"Sec. 3.1, Sec. 3.2, Appendix B, Eq. (6)"},{"comment":"The calibration is built from isolated, uniform-density, 10 K clumps in a 1.53 pc box with no interstellar radiation field, no outflows, and no H II regions. For the stated goal of comparing with real Galactic clumps, these omissions matter. The ISRF experiment in Appendix B shows that external heating strongly changes the inferred mass and over-heats the clump outskirts, and outflows and H II regions are acknowledged in Sec. 4 as future work. The paper should either quantify how these effects would shift the L/M-SFE relation or explicitly restrict the calibration to clumps in which internal heating dominates and feedback is negligible. As written, the conclusion that L/M is 'a reliable parameter' for characterizing the evolutionary stage of observed regions overstates the applicability.","section":"Sec. 2.1, Sec. 4, Appendix B"}],"minor_comments":[{"comment":"The L/M-age fits are quoted with many significant digits, but the zero-points depend on the arbitrary choice of simulation start time as age zero; consider stating the intercept in terms of a reference age or removing it from the text.","section":"Sec. 3.2, Eqs. (8)-(9)"},{"comment":"The text reports mc_weighted_photons=1 for the thermal MC runs and then mc_weighted_photons=0 for the scattering runs; please clarify in one place that these are different settings for the two different Radmc-3d stages.","section":"Sec. 2.2.1"},{"comment":"The caption is very dense and uses gray labels for the parameters of the exact marker; simplifying the legend or splitting it into a table would improve readability.","section":"Fig. 5 caption"},{"comment":"There is a typo in 'There a clear correlation between these two quantities'; it should read 'There is a clear correlation'.","section":"Sec. 3.2"},{"comment":"The statement that the pixel-by-pixel approach 'gives indeed more reliable results' is hard to reconcile with the factor-of-2 overestimate reported for the later stages; please reconcile these statements.","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of A&A and the end-to-end framework is a genuine strength. The main technical shortcoming is the unquantified sensitivity of the calibration zero-point to f_acc and to the SED mass systematics; this is fixable with additional analysis and revised wording. I do not see grounds for rejection, but the current abstract and conclusions claim a quantitative calibration that the error bars do not yet support."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line up front: this is a genuinely useful calibration paper. What is new is the end-to-end pipeline that turns RMHD clump-collapse simulations into Hi-GAL-like synthetic observations and then calibrates L/M against SFE. The slope of 1.20, with mass independence across the explored 500–1000 Msun initial conditions, is a real result, and it is not circular: SFE is sink mass over initial mass, while L/M is reconstructed from synthetic fluxes, and the paper explicitly shows deviations from the direct simulation-side Ltot/Mgas relation.\n\nWhat the paper does well: it replicates observer methods carefully, using Hyper extraction, graybody fits, realistic beams and noise, and it checks the mass reconstruction in Appendix B, reporting 20–40% uncertainties. It also makes an honest point that L/M versus age is mass dependent, while SFE avoids the zero-age problem. The comparison with Paper I's steeper slope of ~1.5 is informative rather than self-serving.\n\nThe soft spot that matters is f_acc. Equation 2 assigns accretion luminosity with an unknown efficiency factor, set to 0.1. In the early and mid SFE regime, accretion luminosity dominates the clump SED, so the entire zero point of Equation 10 follows that choice. A factor-of-2 change moves log(L/M) by about 0.3 dex, and a plausible factor-of-5 change moves the implied SFE at L/M=10 from ~1.3e-2 down to ~3e-3. That directly affects the headline claim that the majority of star formation happens after L/M>10. The quoted ±0.02 slope and ±0.03 intercept errors are pure fit scatter and do not include this. The paper never quantifies or even caveats the f_acc sensitivity. That is a genuine omission, not a manufactured one, and a reviewer should require the sensitivity analysis before the absolute calibration is accepted.\n\nThe other soft spots are smaller and mostly acknowledged. Early-time points are excluded because the imposed 10 K initial temperature corrupts the mass estimate; that is reasonable but means the calibration leans on the later, better-behaved stages. The simulations are isolated uniform-density spheres with no outflows or HII regions, though the paper says this is future work and Appendix B actually tests the ISRF and explains why it was not included. For the slope and mass independence, the idealizations are a concern but not disqualifying.\n\nWho this is for: observers using L/M for Galactic massive clump classification, and simulators who want a template for synthetic observation comparisons. Worth a serious referee. My own verdict would be conditional acceptance: keep the slope claim, but add f_acc and opacity sensitivity tests and soften the absolute SFE translation and the L/M>10 threshold claim until they are done.","headline":"A solid calibration study whose slope is likely robust but whose zero point hinges on one uncalibrated subgrid parameter; referee should require the sensitivity analysis before the absolute L/M-to-SFE translation is trusted.","tokens_in":20624,"tokens_out":2264,"would_cite":true,"duration_ms":30047,"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 calibrates the observed luminosity-to-mass ratio of massive star-forming clumps against star formation efficiency and finds one mass-independent power law, $L/M \\propto {\\rm SFE}^{1.20}$, usable to read evolutionary stage from…","keywords":["star formation efficiency","L/M indicator","massive star-forming clumps","bolometric luminosity","radiative transfer post-processing","synthetic observations","Hi-GAL survey","radiative magnetohydrodynamics"],"falsifier":"Apply Eq. (10) to a sample of real Hi-GAL clumps whose $L/M$ spans roughly 0.5 to 100 $L_\\odot/M_\\odot$, and compare the predicted SFE with independent SFE estimates from ALMA counts of individual protostars (with completeness corrections) or from the mass in detected outflows; a systematic offset—especially for clumps hosting H II regions or showing external heating—would falsify the calibration. A simpler check: clumps with strong outflows should scatter off the power law by more than the quoted scatter if the missing feedback physics matters.","tokens_in":19543,"feed_emoji":"🌟","tokens_out":12983,"duration_ms":113881,"temperature":0.7,"pith_summary":"The paper aims to turn an observable—the bolometric luminosity-to-envelope-mass ratio $L/M$ of parsec-scale star-forming clumps—into a direct readout of how much of a clump's gas has already turned into stars, the star formation efficiency (SFE). From radiative magnetohydrodynamics simulations of collapsing clumps post-processed through synthetic far-infrared observations built to mimic the Hi-GAL survey, the authors reconstruct $L/M$ with the same aperture-photometry and SED-fitting tools observers use, then compare it with the SFE and age recorded in the simulation. They find a power-law correlation $\\log(L_{\\rm bol}/M) = 1.20\\log({\\rm SFE}) + 3.28$ that is independent of clump mass, turbulence level, magnetic-field strength, and turbulent seed. If real Galactic clumps obey the same relation, $L/M$ measured with Herschel-type surveys gives a direct, mass-free estimate of star formation efficiency, and the observed boundary $L/M \\approx 10$ marks the stage past which most star formation still has to occur.","feed_headline":"One power law links clump luminosity to star formation efficiency","feed_subtitle":"Observed L/M translates directly into star formation efficiency, independent of clump mass.","key_machinery":"The engine of the argument is an end-to-end synthetic-observation pipeline. The Ramses adaptive-mesh-refinement code runs radiative magnetohydrodynamics collapse simulations of isolated clumps—uniform-density 10 K spheres of 500 or 1000 $M_\\odot$ in a 1.53 pc box—with sink particles standing in for stars; the Radmc-3d Monte Carlo radiative-transfer code recomputes dust temperatures from stellar and accretion luminosities (accretion efficiency $f_{\\rm acc}=0.1$) using DIANA/MRN dust opacities and ray-traces images at 24, 70, 160, 250, 350, and 500 $\\mu$m; the images are convolved with the real Hi-GAL and MIPSGAL beams and noise; and the Hyper aperture-photometry routine extracts sources whose fluxes are turned into bolometric luminosity and clump mass by SED integration and optically thin graybody fitting. The central identity that carries the result is Eq. (10), the power-law relation converting the observable $L/M$ into the simulation-intrinsic SFE, whose mass independence is the paper's headline claim.","core_discovery":"Stated the way the authors would state it to a fair reader: the observationally reconstructed luminosity-to-mass ratio of massive star-forming clumps is a power-law function of star formation efficiency, $\\log(L_{\\rm bol}/M) = 1.20^{+0.02}_{-0.02}\\log({\\rm SFE}) + 3.28^{+0.03}_{-0.03}$ (Eq. 10), and this relation does not depend on the clump's initial mass or on the initial conditions (turbulent Mach number, magnetic-field strength via $\\mu = 3, 10, 100$, turbulent seed, or viewing projection). By contrast, $L/M$ plotted against clump age is mass-dependent, with separate power laws for the 500 and 1000 $M_\\odot$ realizations, so age cannot be read off without a mass estimate. The paper also maps the three observationally recognized evolutionary phases onto SFE: $L/M \\approx 1$ is reached as soon as the first sink particles form (SFE $\\sim 10^{-3}$–$10^{-2}$), $L/M \\approx 10$ corresponds to SFE $\\sim 10^{-2}$–$5\\times10^{-2}$, and the majority of star formation activity takes place after clumps cross $L/M = 10$.","pith_inferences":["A test the paper does not run: feed real Hi-GAL clump catalogs through Eq. (10) to produce a Galactic SFE distribution, and check whether clumps with similar SFE show similar resolved stellar populations in ALMA follow-ups; agreement would validate the calibration outside the simulated parameter space.","The paper's own Appendix B hints at a systematic direction for real clumps: adding an interstellar radiation field overheats the model clump outskirts, which would bias inferred masses low and hence $L/M$ high for clumps embedded in strong external radiation.","The mass-independence claim rests on a narrow mass window (500–1000 $M_\\odot$); extending the same pipeline to lighter infrared dark clouds and heavier giant clumps could reveal whether the power law bends outside that range, since the $L/M$–age relation already changes steeply with mass.","If the calibration holds, a single far-infrared measurement of $L/M$ gives observers a physically meaningful 'fraction of gas turned into stars', which speaks directly to the clump-fed versus core-fed debate: clumps with low SFE but high $L/M$ would show that a few bright protostars heat the envelope before the bulk of the cluster forms."],"forward_implications":["Observed $L/M$ values from far-infrared surveys translate directly into star formation efficiency for parsec-scale clumps, with no need to know the clump mass.","The three recognized evolutionary phases map onto SFE thresholds: $L/M\\approx1$ appears as soon as the first stars form (SFE $\\sim10^{-3}$–$10^{-2}$) and $L/M\\approx10$ corresponds to SFE $\\sim10^{-2}$–$5\\times10^{-2}$.","Because the SFE is only of order $10^{-2}$ at $L/M=10$, the bulk of a clump's star formation happens after it enters the $L/M>10$ phase.","Absolute clump ages cannot be inferred from $L/M$ without a mass estimate, and strong magnetic fields (mass-to-flux ratio $\\mu=3$) delay star formation by roughly a factor of two.","The post-processed slope (1.20) is shallower than the simulation-intrinsic $L/M$–SFE relation of Paper I (slope $\\simeq1.5$), so the observational mass-reconstruction procedure itself shapes the calibrated power law."],"supporting_citations":[{"why":"Supplies the RMHD simulation suite (RS1.0) whose clumps, sinks, ages, and SFEs are the data the calibration is extracted from.","marker":"Lebreuilly et al. 2025 (Paper I)"},{"why":"Establishes L/M as a three-phase evolutionary indicator for Galactic clumps, the observational scale the paper calibrates.","marker":"Molinari et al. 2016a"},{"why":"Provides the Hyper extraction routine and the SED-integration approach used to reconstruct bolometric luminosities and masses from the synthetic maps.","marker":"Traficante et al. 2015a"},{"why":"Provides the optically thin graybody fitting procedure and the 300-micron opacity reference used to derive dust temperature and clump mass.","marker":"Elia et al. 2017"},{"why":"The Radmc-3d Monte Carlo radiative-transfer code that computes dust temperatures and ray-traced images at the survey wavelengths.","marker":"Dullemond et al. 2012"},{"why":"Supplies the stellar evolution tracks that set sink internal luminosities and radii, feeding the luminosity budget.","marker":"Kuiper & Yorke 2013"},{"why":"The linmix Bayesian regression used to fit and report the power-law slopes with uncertainties.","marker":"Kelly 2007"},{"why":"Defines the Hi-GAL beams and per-band noise rms that make the synthetic observations realistic.","marker":"Molinari et al. 2016b"}],"fun_headline_variants":["L/M power law predicts clump star formation efficiency","Mass-independent law links clump L/M to star formation","Clump evolution clocked by luminosity-to-mass ratio","Star formation efficiency read from clump luminosity ratio"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calibration is built from idealized simulated clumps that start as uniform, isolated, 10 K spheres with no external starlight, no protostellar outflows, and no H II regions; if those missing processes change how dust temperature and inferred envelope mass relate to the true luminosity, real Galactic clumps could fall off the calibrated relation.","fun_headline_variants_meta":{"raw":{"variants":["L/M power law predicts clump star formation efficiency","Mass-independent law links clump L/M to star formation","Clump evolution clocked by luminosity-to-mass ratio","Star formation efficiency read from clump luminosity ratio"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000219,"raw_usage":{"total_tokens":1581,"prompt_tokens":1220,"completion_tokens":361,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":836,"completion_tokens_details":{"reasoning_tokens":298}},"tokens_in":836,"tokens_out":361,"duration_ms":4631,"temperature":1.0,"reasoning_tokens":298,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:44:32.059922+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply Eq. (10) to a sample of real Hi-GAL clumps whose $L/M$ spans roughly 0.5 to 100 $L_\\odot/M_\\odot$, and compare the predicted SFE with independent SFE estimates from ALMA counts of individual protostars (with completeness corrections) or from the mass in detected outflows; a systematic offset—especially for clumps hosting H II regions or showing external heating—would falsify the calibration. A simpler check: clumps with strong outflows should scatter off the power law by more than the quoted scatter if the missing feedback physics matters.","supporting_citations":[{"cited_title":"2017, MNRAS, 471, 100","cited_arxiv_id":null,"evidence_quote":"Provides the optically thin graybody fitting procedure and the 300-micron opacity reference used to derive dust temperature and clump mass."},{"cited_title":"& Yorke, H","cited_arxiv_id":null,"evidence_quote":"Supplies the stellar evolution tracks that set sink internal luminosities and radii, feeding the luminosity budget."}],"review_version":1}