{"id":"36ba568c-d0c7-4171-99d6-d74129fef29a","arxiv_id":"1908.09402","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In 49 merging galaxies with SFR above 1 solar mass per year, the volume and mass of X-ray hot gas scale linearly with the star formation rate, matching stellar feedback simulations.","lead":"Astronomers measured the size and mass of hot X-ray gas in 49 interacting galaxies and found that for the busiest star-forming systems the hot gas mass grows in lockstep with the star formation rate. The result supports computer models in which exploding stars and stellar winds, rather than black holes, are the main engines heating gas in merging galaxies.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Filling-factor f is not measured, so M_X(gas) is really M_true/sqrt(f); if f varies with SFR the claimed linear M_X-SFR and hot/cold mass-ratio slopes are not robust.","rationale":"The reader's weakest_assumption identifies the same issue I find most load-bearing. The central claim is that M_X(gas) is linearly proportional to SFR for SFR > 1 Msun/yr and that M_X(gas)/(M_H2+M_HI) increases with SFR, which is then compared to Moreno et al. simulations. Both statements are about a quantity that contains an undetermined factor 1/sqrt(f). The paper deserves credit for explicitly stating that n_e and f cannot be separated, for quoting the range of plausible f values, and for testing temperature assumptions in Section 6.5. However, it does not re-run any correlation with a varying f, despite arguing in Section 7.2 that f likely increases with SFR. Because the derived M_X is M_true/sqrt(f), a mild f-SFR dependence of slope 0.2-0.4 dex per decade changes the headline slope from 0.88 to roughly 0.98-1.08 or to 0.68-0.78, straddling the linear hypothesis and the flatter alternative. The volume-SFR relation (slope 0.97 +/- 0.15) is not affected by f, so the qualitative conclusion that hot gas production scales with the SFR is secure; the absolute mass scale and the quantitative linearity claim are not. This does not warrant rejection, but it does warrant a conditional verdict pending an explicit f(SFR) sensitivity analysis or independent f constraints.","tokens_in":49681,"tokens_out":6359,"duration_ms":63430,"concrete_test":"Re-run the Section 6.2 and 7.1 regressions with M_X corrected for a bracketing set of filling-factor assumptions: (a) constant f = 0.3, (b) f rising from 0.2 at SFR = 1 Msun/yr to 0.8 at SFR = 100 Msun/yr (a log-linear relation, consistent with the Li et al. 2015 and Breitschwerdt et al. 2012 ranges cited in Section 7.2), and (c) f decreasing from 0.8 to 0.2 over the same range. If the best-fit M_X-SFR slope moves outside 0.88 +/- 0.10, or the M_X/(M_H2+M_HI)-SFR Spearman coefficient drops below the strong/weak boundary, the central claim is not robust to the unknown filling factor. Also report the implied range of absolute M_X values for a few representative galaxies to show the scale uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 5 the paper uses L_X(gas) = Lambda n_e^2 f V and correctly states that only n_e sqrt(f) can be determined, with n_e sqrt(f) between 1.1e-3 and 2.2e-2 cm^-3. The subsequent mass estimate M_X(gas) = m_p n_e V, however, substitutes this quantity for n_e, i.e. it computes m_p V sqrt(L_X/(Lambda V)) = M_true/sqrt(f). The mass scale is therefore only fixed by the implicit choice f = 1; for the f range quoted in Section 7.2 (10-90 percent) the masses are overestimated by factors from 1.05 to 3.2. More importantly, all slope claims built on M_X(gas) absorb an unknown function f(SFR). Since Section 7.2 explicitly argues that f may increase with SFR (citing Li et al. 2015 and others), the derived slope d log M_X(derived)/d log SFR = d log M_true/d log SFR - 0.5 d log f/d log SFR. The headline slope 0.88 +/- 0.10 from Section 7.1 is thus not a measurement of the true M_X-SFR relation unless f is constant or its SFR dependence is known. The same bias enters M_X(gas)/(M_H2+M_HI) ratios and the comparison to Moreno et al. (2019), where the paper's own text acknowledges the degeneracy but does not propagate it into the correlation analysis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses archival Chandra observations of 49 nearby interacting galaxy pairs, mergers, and merger remnants to measure the spatial extent of diffuse hot gas. From the X-ray extent and the thermal luminosity taken from Paper I, the authors derive hot-gas volumes, electron-density estimates, and hot-gas masses, and correlate these with star formation rates, molecular and atomic gas masses, and other galaxy properties. The main claims are that for systems with SFR > 1 Msun/yr the hot-gas volume and mass correlate strongly and near-linearly with SFR, that the ratio of hot gas mass to cold gas mass increases with SFR, and that these trends are consistent with the Moreno et al. (2019) merger simulations. The paper also reports weak anti-correlations between M_X(gas)/SFR and dust temperature tracers and identifies possible excess hot gas in low-SFR, high-stellar-mass systems such as NGC 1700.","tokens_in":50027,"tokens_out":3221,"duration_ms":37471,"significance":"If the derived masses are reliable, the paper provides a valuable observational constraint on stellar feedback in merging systems using independent X-ray, CO, and HI data. Its strengths include a well-documented procedure for measuring X-ray extents, explicit checks against an alternative extraction method, tests of the assumed gas temperature, and comparisons using two CO-to-H2 conversion schemes. The comparison to an external simulation (Moreno et al. 2019) is a genuine test rather than a fit. The central correlation between hot-gas volume and SFR does not depend on the filling factor and appears robust. However, the hot-gas mass scale and all mass-based slopes depend on the unknown filling factor, as discussed below, so the headline mass-SFR slope and the hot-to-cold mass ratio comparison should be interpreted with caution until that degeneracy is addressed.","major_comments":[{"comment":"The derivation of M_X(gas) implicitly sets the filling factor f = 1. In Section 5 the paper correctly states that L_X(gas) = Lambda n_e^2 f V and that only the product n_e sqrt(f) can be determined, but then computes M_X(gas) = m_p n_e V by substituting the measured quantity n_e sqrt(f) for n_e. The resulting quantity is m_p V sqrt(L_X/(Lambda V)) = M_true/sqrt(f). For the range f ~ 0.1-0.9 quoted in Section 7.2, the masses are overestimated by factors of 1.05-3.2. More importantly, if f varies with SFR, then the derived slope d log M_X(derived)/d log SFR equals d log M_true/d log SFR - 0.5 d log f/d log SFR. Since Section 7.2 explicitly notes that simulations predict f to increase with the density of star formation and hence potentially with SFR, the headline slope of 0.88 +/- 0.10 in Section 7.1 is not a measurement of the true M_X-SFR relation unless f is constant or its SFR dependence is known. The authors should re-run the correlation analysis with a parameterized f(SFR) (e.g., using the Li et al. 2015 range) or otherwise quantify how the slope shifts.","section":"Section 5 and Section 7.1"},{"comment":"The same filling-factor degeneracy propagates into the hot-to-cold gas mass ratio M_X(gas)/(M_H2 + M_HI) and the comparison with the Moreno et al. (2019) simulations. If f increases with SFR while the true M_X-SFR slope is shallower than the derived slope, the reported increase of M_X(gas)/(M_H2 + M_HI) with SFR (slope 0.82 +/- 0.16 for the variable CO/H2 ratio and SFR > 1 Msun/yr, Table 5) could be partly an artifact of the 1/sqrt(f) factor. The paper acknowledges the degeneracy in Section 7.2 but does not propagate it into the correlation analysis or the simulation comparison. A quantitative test with an assumed filling-factor model is needed before the agreement with Moreno et al. can be considered established.","section":"Sections 6.3 and 7.1"}],"minor_comments":[{"comment":"The figures showing the key correlations omit error bars entirely, despite the paper quoting a factor-of-two uncertainty in M_X(gas) and comparable uncertainties in volume. At minimum, representative error bars should be shown on the most load-bearing plots (Figures 13 and 15) so readers can judge the scatter against the uncertainties.","section":"Figures 9-19"},{"comment":"Several rows in Table 5 refer to 'LOG n_e' (e.g., in the comparisons with volume and with M_X(gas)/SFR), but the paper only measures n_e sqrt(f), not n_e alone. The column headers and entries should be relabeled as log(n_e sqrt(f)) to avoid implying an independent determination of n_e.","section":"Table 5"},{"comment":"In the paragraph beginning 'As with NGC 1700', the text reads 'separating out this additional component to the hot gas is is uncertain'; the duplicated 'is' should be removed.","section":"Section 6.2"},{"comment":"The summary states 'we see a possible deficient of hot gas in low mass systems'; 'deficient' should be 'deficiency'.","section":"Section 8"},{"comment":"In the list of parameters that are not correlated, the sentence 'M X(gas)/SFR is not correlated with for a variable CO/H 2 ratio' is missing its subject; it should presumably read 'not correlated with SFE for a variable CO/H2 ratio'.","section":"Section 7.2"}],"recommendation":"major_revision","confidential_remarks":"The filling-factor issue is acknowledged by the authors in Section 7.2 but is central enough to the mass-SFR slope and the simulation comparison that it should be addressed quantitatively before publication. A relatively small addition—recomputing the mass-based correlations under an assumed f(SFR) relation or at least presenting the slope as a function of f—would substantially strengthen the paper. I see no scope or novelty concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague—\n\nThe short version: this is a careful archival paper with genuinely new measurements, and its main qualitative claims probably survive the one real flaw in the mass estimate. The stress-test note is right, but the paper is not broken.\n\nWhat's new: Paper I gave L_X-SFR. This one measures the spatial extent of hot gas in 49 mergers, derives volumes and electron densities, and then compares M_X(gas) to cold gas masses from published CO and HI. The relation between hot gas volume and SFR is strong and independent of filling factor. The trend of increasing M_X(gas)/(M_H2+M_HI) with SFR and the comparison to Moreno et al. (2019) are new. The authors test alternate temperatures and two CO conversion factors, and they flag their own weak statistics.\n\nThe soft spot is exactly where the stress-test points. In Section 5 they define L_X = Lambda n_e^2 f V, correctly note that only n_e sqrt(f) is measured, then compute M_X = m_p n_e V as if f=1. The absolute masses are overestimated by 1/sqrt(f), and if f increases with SFR, the derived slope 0.88 ± 0.10 is not the true slope. Since Section 7.2 itself argues f may vary with SFR, the linear M_X-SFR claim is not quantitatively nailed down. The hot-to-cold ratios have the same bias. I don't think this kills the paper, because the volume-SFR correlation is f-free and the qualitative M_X-SFR trend is very unlikely to vanish for any plausible f(SFR), but the specific slope and the mass scale should be treated as conditional.\n\nMinor complaints: no error bars on the plots (they state factor-of-two uncertainties); CO data incomplete; sample is archive-selected. None of these are hidden; the authors are candid about them.\n\nWho should read it: observers and simulators working on feedback in mergers. It's a useful data point for the subfield, not a paradigm change. Yes, it deserves a serious referee. I'd accept it with moderate revision, mainly asking them to present the f-dependence of their slopes explicitly, and to add error bars where feasible.","headline":"New hot-gas volume and mass measurements for 49 mergers; the M_X-SFR slope is real but only conditional because the mass estimate implicitly sets the filling factor to unity.","tokens_in":50600,"tokens_out":2758,"would_cite":true,"duration_ms":28198,"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":"In merging galaxies, hot gas mass tracks star formation rate linearly.","keywords":["hot gas","X-ray emission","star formation rate","galaxy mergers","stellar feedback","interstellar medium","Chandra","molecular gas"],"falsifier":"Measure electron densities independently of the filling factor for a subsample of these systems — for example, from X-ray line ratios, absorption measurements, or dispersion toward background sources — and recompute M_X(gas) using the measured f values; if f varies systematically with SFR, the claimed unity-slope M_X-SFR relation should either survive with f included or flatten, which would show that the linear relation was an artifact of the constant-f assumption.","tokens_in":49482,"feed_emoji":"🔭","tokens_out":6577,"duration_ms":65247,"temperature":0.7,"pith_summary":"Using Chandra archival imaging of 49 nearby interacting galaxies, pairs, mergers, and remnants, this paper measures the spatial extent of diffuse soft X-ray emission and derives volumes, electron densities, and masses of the hot interstellar gas. It claims that for systems forming stars faster than about one solar mass per year, both the volume and the mass of hot gas are nearly proportional to the star formation rate, with log-log slopes of 0.97 ± 0.15 for volume and 0.88 ± 0.10 for mass. That linearity is the signature the authors read as stellar winds and supernovae, rather than older stars, setting the hot-gas budget during a starburst, and it agrees with recent merger simulations in which the ratio of hot to cold gas rises with star formation. The paper also finds that the hot-to-cold gas mass ratio increases with star formation rate and with dust temperature, while the hot gas mass per unit star formation falls as the 60-to-100 micron flux ratio rises.","feed_headline":"Hot gas mass tracks star formation rate in galaxy mergers","feed_subtitle":"Chandra data on 49 interacting galaxies show supernova feedback, not old stars, sets the hot-gas budget during bursts.","key_machinery":"The load-bearing identity is the X-ray emission relation $L_X(\\mathrm{gas}) = \\Lambda n_e^2 f V$, relating the measured thermal luminosity to the cooling function $\\Lambda$, the electron density $n_e$, the volume filling factor $f$, and the hot-gas volume $V$. Since only the product $n_e\\sqrt{f}$ is determined, the paper sets the hot-gas mass to $M_X(\\mathrm{gas}) = m_p n_e V$, effectively assuming $f=1$ and overestimating the mass by $1/\\sqrt{f}$ if the gas is clumpy. Volumes come from ellipses fitted to 0.3–1.0 keV maps at a common surface-brightness cutoff of $3\\times10^{-9}$ photons s$^{-1}$ cm$^{-2}$ arcsec$^{-2}$, chosen so that the enclosed counts match those inside the optical isophote within about 10 percent. This machinery converts the X-ray images plus spectra into the correlated quantities — hot gas volume, hot gas mass, and the hot-to-cold gas ratio — that are then compared with SFR, the stellar mass proxy $L_K$, dust temperature, and star-formation efficiency.","core_discovery":"The central discovery claim is that in these 49 merging systems the hot X-ray-emitting gas behaves like a direct exhaust product of the current starburst: for SFR $> 1\\,M_\\odot\\,\\mathrm{yr}^{-1}$, $\\log M_X(\\mathrm{gas})$ versus $\\log\\,\\mathrm{SFR}$ has slope $0.88 \\pm 0.10$, and the hot gas volume versus SFR has slope $0.97 \\pm 0.15$, both consistent with proportionality. The mass is computed from $L_X = \\Lambda n_e^2 f V$ under an assumed temperature of 0.3 keV where no spectral temperature is available, and the hot-to-cold ratio $M_X(\\mathrm{gas})/(M_{\\mathrm{H}_2}+M_{\\mathrm{HI}})$ increases with SFR, with a slope consistent with unity when a variable CO-to-H$_2$ ratio is used and low-SFR systems are excluded. This is presented as support for stellar and supernova feedback as the dominant hot-gas source in starbursting mergers, and as an observational match to hydrodynamic simulation predictions. Additional correlations, including an excess of hot gas in low-SFR, high-stellar-mass remnants and a possible deficit in low-mass systems, are flagged as uncertain because the sample contains few such galaxies.","pith_inferences":["If the linear $M_X$ – SFR relation is real, it supplies a direct conversion from an easily measured SFR to the size of the hot reservoir that regulates future star formation; one consequence is that the hot-gas mass could serve as a feedback calibrator for galaxy simulations without needing to resolve individual supernovae.","The filling-factor degeneracy means the published masses are upper limits; if $f$ varies systematically with SFR, as simulations suggest for higher-density star formation, the reported slope could steepen or flatten once real filling factors are included, so an independent electron-density measurement is the decisive test.","Applying the same surface-brightness-cutoff method to isolated starbursts and post-starbursts, not just mergers, would separate merger-specific effects such as tidal compression and triggered bursts from the universal feedback relation claimed here.","If the low-SFR excess is indeed old-stellar mass loss, then the hot-gas mass of a post-merger elliptical may act as a clock: comparing $M_X/\\mathrm{SFR}$ across remnants of different ages could map how the virialized hot halo accumulates after the starburst fades."],"forward_implications":["For high-SFR systems, the near-unity slopes imply that every unit of star formation produces roughly a fixed amount and volume of hot gas, so a merger boosts the hot-gas reservoir simply by boosting the star formation rate.","The rising hot-to-cold gas mass ratio with SFR, consistent with hydrodynamic merger simulations, means the starburst redistributes interstellar mass toward the hot phase without destroying the cold reservoir that sustains the burst.","The constancy of $M_X(\\mathrm{gas})/\\mathrm{SFR}$ at high SFR, with scatter near the measurement uncertainty of about 0.34–0.37 dex, argues for a near-steady-state feedback loop operating on a timescale comparable to the radiative cooling time and the roughly 100 Myr averaging time of the SFR indicator.","Low-SFR, high-stellar-mass merger remnants deviate upward from the $M_X$ – SFR relation, identifying a separate hot-gas channel — likely virialized mass loss from old stars — that dominates when star formation is weak.","The mild anti-correlation of $M_X(\\mathrm{gas})/\\mathrm{SFR}$ with the 60/100 micron flux ratio and with the 3.6–24 micron color points to the spatial concentration of young stars or the efficiency of early feedback, not the total SFR, as a second driver of hot-gas yield per unit star formation."],"supporting_citations":[{"why":"Parent sample definition; supplies the diffuse MEKAL X-ray luminosities, distances, and UV-plus-IR star formation rates used throughout this paper.","marker":"Smith et al. 2018"},{"why":"Hydrodynamic merger simulations with stellar feedback that predict the hot-to-cold gas mass ratio increases during starbursts; this is the key comparison target for the observed trend.","marker":"Moreno et al. 2019"},{"why":"Establishes the L_X(gas)-SFR proportionality for star-forming galaxies and provides comparison electron densities for the derived n_e sqrt(f) values.","marker":"Mineo et al. 2012b"},{"why":"Provides the method of measuring hot gas radial extent via an enclosed-light surface brightness level, adapted here to a fixed 3e-9 photons s^-1 cm^-2 arcsec^-2 cutoff.","marker":"Strickland et al. 2004a"},{"why":"Earlier L_X(gas)-SFR study whose adopted hot-gas extent and feedback energy fraction serve as calibration points for the current analysis.","marker":"Grimes et al. 2005"},{"why":"Theoretical framework attributing hot gas to stellar winds, radiation pressure, and supernovae, used to interpret the linear M_X-SFR relation.","marker":"Hopkins et al. 2012a"},{"why":"Provides the SFR calibrations and the roughly 100 Myr averaging timescale for the UV-and-IR SFR estimates.","marker":"Kennicutt & Evans 2012"}],"fun_headline_variants":["Hot gas in mergers scales with star formation rate","Starbursts blow hot gas: Chandra survey finds direct link","Merger hot gas tracks SFR, boosting feedback models","Chandra: hot gas mass follows star formation in 49 mergers","Supernova feedback sets hot gas budget in galaxy mergers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The hot-gas mass is calculated as if the emitting gas fills the entire fitted volume; only the product of the electron density and the square root of the filling factor is actually measured, so the quoted masses and all correlations built on them assume that the filling factor is either unity or does not vary systematically with star formation rate.","fun_headline_variants_meta":{"raw":{"variants":["Hot gas in mergers scales with star formation rate","Starbursts blow hot gas: Chandra survey finds direct link","Merger hot gas tracks SFR, boosting feedback models","Chandra: hot gas mass follows star formation in 49 mergers","Supernova feedback sets hot gas budget in galaxy mergers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000404,"raw_usage":{"total_tokens":2211,"prompt_tokens":1162,"completion_tokens":1049,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":778,"completion_tokens_details":{"reasoning_tokens":967}},"tokens_in":778,"tokens_out":1049,"duration_ms":8427,"temperature":1.0,"reasoning_tokens":967,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:12:36.376078+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure electron densities independently of the filling factor for a subsample of these systems — for example, from X-ray line ratios, absorption measurements, or dispersion toward background sources — and recompute M_X(gas) using the measured f values; if f varies systematically with SFR, the claimed unity-slope M_X-SFR relation should either survive with f included or flatten, which would show that the linear relation was an artifact of the constant-f assumption.","supporting_citations":[{"cited_title":"J., Campbell, K., Struck, C., et al.\\ 2018, AJ, 155, 81","cited_arxiv_id":null,"evidence_quote":"Parent sample definition; supplies the diffuse MEKAL X-ray luminosities, distances, and UV-plus-IR star formation rates used throughout this paper."},{"cited_title":"L., et al.\\ 2019, MNRAS, 485, 1320","cited_arxiv_id":null,"evidence_quote":"Hydrodynamic merger simulations with stellar feedback that predict the hot-to-cold gas mass ratio increases during starbursts; this is the key comparison target for the observed trend."}],"review_version":1}