{"id":"f2b5f0e7-63e0-40fd-9073-0aa189d5f170","arxiv_id":"1908.05274","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Gas rotation curves in simulated high-redshift galaxies underestimate dynamical mass by up to 40% unless turbulent pressure support and non-spherical potentials are corrected.","lead":"Using cosmological simulations, this paper shows that gas rotation alone underestimates the masses of high-redshift galaxies, by up to 40% in the outer disks, unless turbulent pressure and the shape of the gravitational potential are properly corrected. It is a calibration check for the telescope surveys that measure galaxy masses from gas motion.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Mass recovery in Fig. 6 is an algebraic identity under the circularity cuts, while the observational pressure proxy tested in Fig. 7 under-predicts support; the central claim's transfer to real data is therefore not established.","rationale":"The reader's weakest assumption was that FIRE-2 faithfully reproduces the ISM turbulence of real high-redshift galaxies, so that the quantitative 40% bias transfers to observations. That is a legitimate concern, explicitly flagged by the authors in Section 4.2. My stress-test identifies a more immediate and internal problem: the recovery shown in Figures 4 and 6 uses the full 3D gravitational force and 3D pressure gradient, which are unavailable in real observations, and Eq. (9) is an identity given the momentum equation and the definitions of the correction terms. The paper itself shows in Figure 7 that the observational proxies one can actually build from Σ and line-of-sight σ under-predict the pressure support in the disk body. Therefore the Section 5 statement that 'the total mass profile may be successfully recovered' has not been demonstrated for observable quantities. This does not overturn the paper's central physical insight — pressure gradients matter and the standard proxy fails — but it changes the epistemic status of the headline. A forward-modeled mock-observation test would settle whether an observer using standard fitting tools plus the proposed corrections can recover Menc. The reader's CONDITIONAL verdict already captures this uncertainty, so I do not move the verdict. I partially agree with the reader because the identity/proxy gap is related to but distinct from the external FIRE fidelity question: even if FIRE-2 is perfect, the recovery claim would still lack an observational closure test; conversely, if the mock-observation test passes, the external turbulence caveat would remain the main residual uncertainty.","tokens_in":23497,"tokens_out":5658,"duration_ms":62708,"concrete_test":"Forward-model the FIRE-2 massive disk snapshots as synthetic ALMA/IFU observations: project the z ≈ 1.5-2 snapshots that pass the disk criteria at several inclinations, convolve to realistic angular resolution (e.g., 0.1-0.3 arcsec), add noise, and fit the cubes with a standard disk-fitting code such as 3D-Barolo or GalPak3D to obtain observed v_phi(r), σ_1d(r), and Σ(r). Apply the published asymmetric-drift correction of Eq. (11) with both the double-power-law and exponential surface-density options, and compare the resulting M_est(r) with the true Menc(r) in the simulations. If median M_est/Menc differs from unity by more than ~10% in the 1-3 kpc disk body, the paper's central recovery claim does not hold for observable data and the abstract's 40% bias figure should be recast as a simulation-specific upper bound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 5 and the abstract is that with a well-organized disk, circular orbits, and accounting for turbulent pressure and non-spherical potentials, 'the total mass profile may be successfully recovered' from gas rotation. The demonstration of this claim in Figure 6 is weaker than it appears. Equation (9) is not an empirical test but an algebraic rearrangement of the radial momentum equation using simulation-internal quantities: with f_g defined as GMenc/r^2 + δf_g (Eq. 3) and P = ρσ^2, substituting Eq. (8) into Eq. (4) yields Menc = M_vphi + ΔMgrav + ΔMpress identically. The only physical content is that advective terms and non-axisymmetric forces are negligible in the selected annuli; the closeness of the purple curve to unity therefore mostly validates the circularity/equilibrium selection criteria in Section 2.2.4, not the observational feasibility of the corrections. In actual observations, f_g and d log(ρσ^2)/d log r are not directly measured; one instead uses proxies such as Eq. (11), and Figure 7 shows all three quasi-observational estimators systematically under-predict the pressure term in the main disk body (with small-aperture σ measurements making this worse). Thus the paper's own Figure 7 undercuts the Section 5 recovery claim if 'recovered' means recovered from observables. What is robustly demonstrated is that turbulent pressure gradients can be a large physical effect (10-40% mass bias in these simulations) and that the standard surface-density proxy fails in the disk body. The additional step from these demonstrations to 'total mass profile may be successfully recovered' requires an observable-space closure test that the paper does not provide.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses four high-resolution FIRE-2 cosmological zoom-in simulations of massive z=1-3 galaxies, plus three lower-mass Milky-Way-mass runs, to assess the corrections required when inferring enclosed dynamical masses from gas rotation curves. The authors measure the intrinsic azimuthal velocity, the radial turbulent pressure gradient, the non-spherical gravitational force, and the degree of orbital circularity in the cool gas, and show that after applying the pressure-gradient and aspherical-potential corrections in the circularized regions of disk snapshots, the summed mass estimate matches the true enclosed mass (Fig. 6). They also test commonly used observational proxies for the pressure term (Eq. 11) and find that these systematically underpredict the true pressure support in the main body of the disk (Fig. 7). The paper concludes that turbulent pressure gradients can bias dynamical mass estimates low by up to 40% in the outer disk and that the full mass profile can be recovered if the corrections are properly included, with the caveat that the simulated galaxies are over-compact and more turbulent than typical observed systems.","tokens_in":23739,"tokens_out":9573,"duration_ms":96690,"significance":"The paper's direct, simulation-internal measurement of the turbulent-pressure and nonspherical-potential corrections is a strength, and its demonstration that these corrections close the radial force balance in the selected annuli is convincing. The central physical result—that pressure gradients can lower inferred masses by tens of percent in the outer, steep-density regions of high-redshift disks—is important and well supported, and the analysis is careful about the disk and circularity selection. The authors are also transparent about the known limitations of the simulations (missing AGN feedback, over-compact stellar distributions, elevated velocity dispersions) and provide a data/code availability statement. The significance is somewhat tempered by the fact that the paper does not demonstrate that these corrections can be measured accurately from observable quantities; indeed, its own Figure 7 shows that standard observational proxies fail to capture the pressure support in the main disk body. Nevertheless, the work is a useful step toward quantifying a systematic that has been identified in observational rotation-curve analyses.","major_comments":[{"comment":"The recovery claim in Section 5 ('the total mass profile may be successfully recovered') is substantially stronger than what the tests demonstrate. Equation (9) is a rearrangement of the radial momentum equation: with f_g defined by Eq. (3) and P = rho sigma^2, Eq. (9) holds by construction for any annulus in which the neglected advective and non-axisymmetric forces are small. The closeness of the purple curve to unity in Figure 6 is therefore primarily a check of the circularity criteria in Section 2.2.4, not a demonstration that a realistic observer can recover M_enc. That latter point is directly challenged by Figure 7, which shows that all three quasi-observational estimators of the pressure term (Eq. 11) underpredict Delta M_press in the main body of the disk. To support the Section 5 claim, the paper needs either to rephrase 'recovered' to make clear that the correction uses simulation-truth forces, or to add a mock-observation analysis that propagates the proxy errors and shows the resulting mass estimates.","section":"Section 5, Section 4.2, Figure 7"},{"comment":"The headline quantitative result—'bias dynamical mass measurements low by up to 40%'—is tied to the simulated ISM velocity dispersions of 100-150 km/s, which the paper itself notes are higher than typically expected from observations. Because the pressure-support term scales approximately as sigma^2/r, a real galaxy population with sigma of order 50-70 km/s would have a substantially smaller bias. Please add an explicit sensitivity estimate (e.g., rescaling sigma to observed values or an analytic scaling argument) and move the caveat into the abstract so that the 'up to 40%' figure is not read as a universal prediction.","section":"Section 4.2, abstract"},{"comment":"The paper assumes that the non-circular regions can be 'identified and excluded' in observations, citing Oman et al. (2019), but does not demonstrate that this selection can be made reliably at the spatial resolution and signal-to-noise of high-redshift IFU or ALMA data. Since the recovery in Figure 6 applies only to the selected circularized region, the practical recommendation for observers is incomplete without a test of how well the selection works on realistic mock observations.","section":"Section 4.3"}],"minor_comments":[{"comment":"The text defines P(r) as a sum over particle masses, but it is not immediately clear that this is a volume-averaged pressure; please state explicitly that P(r) = rho sigma_vr^2 with the volume element 2 pi r Delta r times 2 z_h(r), and that sigma_vr is the one-dimensional radial velocity dispersion used throughout.","section":"Section 2.2.2"},{"comment":"The reference list contains a duplicate entry for El-Badry et al. (2018); please merge the two entries.","section":"References"},{"comment":"The shaded regions represent the 25th-75th percentile of the variation among snapshots, but the number of snapshots contributing to each panel is not given; please report these numbers, particularly for the z=2.5-3 bins where A4 and A8 have few or no disky snapshots.","section":"Figure 6"},{"comment":"The approximation 'approx -2 sigma^2 d log Sigma / d log r' is stated without derivation; a footnote spelling out the isothermal-sheet assumption (rho_0 proportional to Sigma^2, constant scale height) would help readers assess the proxy's validity.","section":"Equation (11)"}],"recommendation":"major_revision","confidential_remarks":"The paper makes a solid, direct measurement of a physically important systematic, but the public abstract overstates the observational transfer of the recovery claim. The two main issues—the algebraic-identity nature of the recovery test and the failure of the tested observational pressure proxies—are fixable by rewording and by adding sensitivity tests; no new simulations are required. The paper is within the scope of MNRAS and would be a useful addition after these revisions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing worth knowing about this paper is that it measures, directly inside FIRE-2, how much turbulent pressure gradients can bias gas-based dynamical masses. In the outer disks of their massive high-z galaxies the effect is 10-40%, and it changes the shape of the rotation curve, not just the normalization. The second useful result—the one observers should actually care about—is that the standard analytic proxy, 2σ² dlogΣ/dlogr, systematically under-predicts the true pressure support in the main body of the disk, while working better in the outer disk. That is a concrete, falsifiable statement about a widely used correction.\n\nThe paper does this carefully. They compute the gravitational force by brute force from the simulation, measure the pressure gradient from the full 3D data, and are transparent about the main caveats: no AGN feedback, over-compact stellar distributions, velocity dispersions higher than typical observed values. The Milky-Way-mass appendix helps show the effects are not unique to their extreme massive galaxies.\n\nWhere it gets soft: the claim that 'the total mass profile may be successfully recovered' is not really demonstrated. Equation (9) is the radial momentum equation rewritten as a sum of mass terms. If you measure all three terms exactly from the simulation, they have to sum to M_enc. Figure 6 therefore mostly confirms that their circularity cuts select regions where advection and non-axisymmetric forces are small—that is a valid consistency check, but it is not a test of whether an observer, working from observable quantities, could apply these corrections. The real test is Figure 7, and there all three quasi-observational estimators miss the true pressure support in the disk body. So the abstract overstates what is shown: the paper establishes that the physical effect is large and that the standard proxy fails, but not that the corrections can be implemented observationally.\n\nThe 40% number also carries more uncertainty than the abstract implies. These simulated galaxies are more compact and more turbulent than observed ones, and the paper says so. If real z>1 disks are less extreme, the bias is smaller. That is a limitation, but not a fatal one—the qualitative conclusion that pressure gradients matter and that surface-density-based proxies are unreliable in the disk body should survive.\n\nWho gets value: anyone interpreting IFU or ALMA rotation curves at z>1, especially people relying on the Genzel et al. (2017) asymmetric-drift correction. This paper should be read, and cited, for the proxy failure result. The recovery claim should be reworded and accompanied by an observable-space closure test or mock observations.\n\nFor peer review: yes, a serious referee should look at this. The math is transparent, the simulations are state of the art, and the negative result about the proxy is important even if the paper needs revision on what 'recovered' means.","headline":"The genuinely useful result is that the standard 2σ²dlogΣ/dlogr proxy under-predicts turbulent pressure support in the disk body of FIRE-2 galaxies, but the paper's 'successful recovery' of the mass profile is largely an identity rather than an observational demonstration.","tokens_in":24496,"tokens_out":1963,"would_cite":true,"duration_ms":21048,"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":"Gas rotation curves in high-redshift disk galaxies can recover the true enclosed-mass profile once turbulent pressure support, non-spherical potentials, and non-circular orbits are accounted for, and neglecting the pressure term alone…","keywords":["galaxy kinematics","dynamical masses","high-redshift galaxies","turbulent pressure support","asymmetric drift","rotation curves","cosmological zoom-in simulations","interstellar medium turbulence"],"falsifier":"A decisive test is to compare pressure-corrected masses from rotation curves with independent strong-lensing masses for the same $z\\approx1$--$2$ galaxies; systematic residuals that grow with radius, or a failure of the correction in galaxies with velocity dispersions well below 100 km/s, would show that the simulated pressure structure does not transfer to real disks.","tokens_in":23181,"feed_emoji":"🌀","tokens_out":10181,"duration_ms":95607,"temperature":0.7,"pith_summary":"This paper argues that rotation curves of cool gas in high-redshift disk galaxies can be converted into accurate enclosed-mass profiles, provided the gas sits in a smooth, rotation-dominated disk on nearly circular orbits and the radial gradient of turbulent pressure is added back. In the paper's high-resolution cosmological zoom-in simulations, gas rotation matches the circular velocity $\\sqrt{GM_{\\rm enc}/r}$ in an intermediate radial band but falls below it in the outer disk. The shortfall is largely turbulent pressure support; neglecting it biases dynamical mass estimates low by up to 40 percent. The paper also finds that observational proxies that estimate the pressure term from surface-density gradients typically under-correct inside the disk, and that with all corrections combined the total mass profile is recovered on average.","feed_headline":"Gas turbulence can hide 40% of a galaxy's mass","feed_subtitle":"In simulated high-redshift disks, neglecting turbulent pressure support biases dynamical masses low by up to 40%.","key_machinery":"The paper's machinery is a term-by-term decomposition of the enclosed mass into the mass implied by rotation alone, a gravitational correction for non-spherical potentials, and a pressure correction for radial turbulent-pressure gradients: $M_{\\rm enc} = M_{\\bar v_\\phi} + \\Delta M_{\\rm grav} + \\Delta M_{\\rm press}$, with $\\Delta M_{\\rm press} = -\\frac{\\sigma^2 r}{G}\\frac{d\\log(\\rho\\sigma^2)}{d\\log r}$. In the simulations each term is measured directly from particle data--brute-force gravitational accelerations on test particles and the pressure profile of cool gas--so the paper can test which physical effects actually account for the discrepancy between $\\bar v_\\phi$ and $\\sqrt{GM_{\\rm enc}/r}$. The machinery also includes selection cuts requiring rotation dominance, surface-density smoothness, and orbital circularity (small radial inflow and azimuthally uniform radial motion) to identify the region where rotation can be interpreted as circular motion.","core_discovery":"The central claim is that the full set of dynamical corrections--turbulent pressure gradients, a non-spherical gravitational potential, and exclusion of non-circular orbital regions--accounts for essentially all of the difference between measured gas rotation and the spherical, zero-pressure expectation in simulated disks. Concretely, the paper derives and tests a mass decomposition $M_{\\rm enc} = M_{\\bar v_\\phi} + \\Delta M_{\\rm grav} + \\Delta M_{\\rm press}$ and shows that the corrected sum matches the true enclosed mass on average once only disk snapshots and circularized annuli are selected. The pressure term alone reduces the inferred mass by 10--40 percent in the outer disk, where the surface-density profile steepens, and can exceed 10 percent throughout the disk at $z>2$. The non-spherical-potential correction matters mainly inside the central kiloparsec. The paper concludes that with these conditions met, the total mass profile may be successfully recovered from gas rotation.","pith_inferences":["Extension: if real high-redshift disks are typically less turbulent than the simulated 100--150 km/s dispersions, the 40 percent figure is an upper bound, and the size of the correction should correlate with measured velocity dispersion in a way that can be tested directly.","Extension: kinematic samples that trace only bright clumps or CO peaks, where only small-aperture velocity dispersions are available, would systematically under-correct for pressure and therefore keep inferred masses biased low.","Extension: the paper's circularity criteria suggest a practical quality flag--galaxies with quadrant-to-quadrant variation in radial velocity above roughly 100 km/s, or with azimuthal velocity less than 90 percent of the in-plane speed, should be flagged as unsafe for dynamical mass inference."],"forward_implications":["Neglecting the turbulent-pressure gradient biases dynamical masses of high-redshift disks low by 10--40 percent in the outer disk, so published masses that omit this correction are systematic underestimates there.","Outer rotation-curve declines in high-redshift galaxies can be produced by pressure support rather than by a falling circular velocity, so they should not be read directly as evidence of low dark-matter content.","Standard quasi-observational estimates of pressure support from surface-density slopes (e.g., $-2\\sigma^2\\,d\\log\\Sigma/d\\log r$) under-correct in the main disk because the surface-density profile is shallow; observers should use steeper outer-disk prescriptions and avoid measuring $\\sigma$ in small azimuthal segments.","At $z>2$, pressure support can bias mass estimates low throughout the disk, so high-redshift kinematic samples need the correction even when they do not reach the far outskirts.","The same analysis applied to Milky-Way-mass simulated disks at $z\\approx0$ shows the correction is small except in the outer disk, supporting the standard local practice of treating rotation as tracing the potential."],"supporting_citations":[{"why":"Supplies the galaxy-formation model used to run the simulated galaxies analyzed throughout the paper.","marker":"Hopkins et al. 2018"},{"why":"Provides the four massive high-redshift zoom-in halos used for the main analysis.","marker":"Anglés-Alcázar et al. 2017b"},{"why":"Introduces the turbulent-pressure (asymmetric-drift) correction that the paper measures and tests against observable proxies.","marker":"Burkert et al. 2010"},{"why":"The observational high-redshift kinematics study whose declining rotation curves and pressure-corrected dark-matter fractions motivate the analysis.","marker":"Genzel et al. 2017"},{"why":"An observational sample whose inferred dynamical masses fall below baryonic masses, which the paper's corrections would address.","marker":"Price et al. 2019"}],"fun_headline_variants":["Turbulent pressure biases galaxy mass estimates by up to 40%","Correcting gas kinematics reveals true galaxy masses","Simulations: gas turbulence skews mass measurements","Pressure support distorts mass measurements in high-z disks","Gas rotation underestimates mass without pressure correction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative case rests on the paper's cosmological zoom-in simulations faithfully reproducing the interstellar-medium turbulence of real high-redshift disks, because its velocity dispersions of roughly 100--150 km/s are higher than typical observed values and would inflate the pressure-gradient bias.","fun_headline_variants_meta":{"raw":{"variants":["Turbulent pressure biases galaxy mass estimates by up to 40%","Correcting gas kinematics reveals true galaxy masses","Simulations: gas turbulence skews mass measurements","Pressure support distorts mass measurements in high-z disks","Gas rotation underestimates mass without pressure correction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000175,"raw_usage":{"total_tokens":1346,"prompt_tokens":1065,"completion_tokens":281,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":681,"completion_tokens_details":{"reasoning_tokens":205}},"tokens_in":681,"tokens_out":281,"duration_ms":3388,"temperature":1.0,"reasoning_tokens":205,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:19:17.657157+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to compare pressure-corrected masses from rotation curves with independent strong-lensing masses for the same $z\\approx1$--$2$ galaxies; systematic residuals that grow with radius, or a failure of the correction in galaxies with velocity dispersions well below 100 km/s, would show that the simulated pressure structure does not transfer to real disks.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the galaxy-formation model used to run the simulated galaxies analyzed throughout the paper."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the turbulent-pressure (asymmetric-drift) correction that the paper measures and tests against observable proxies."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The observational high-redshift kinematics study whose declining rotation curves and pressure-corrected dark-matter fractions motivate the analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"An observational sample whose inferred dynamical masses fall below baryonic masses, which the paper's corrections would address."}],"review_version":1}