{"id":"1d62335d-f6d3-4c05-8d69-cff2ff0e357a","arxiv_id":"2506.17554","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Turbulence determines where carbon ions sit in simulated galaxy halos, while background ultraviolet radiation sets the overall ionization balance, boosting C II and C IV.","lead":"This study uses 3D simulations of turbulent gas around galaxies to show that turbulence and background radiation play different roles in shaping carbon ions. Turbulence makes the gas clumpy and shock-heated, while radiation shifts most carbon into C II and C IV, helping explain what telescopes may see in the gas around galaxies.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Radiation changes C IV spatial clustering as much as turbulence does, undercutting the claimed spatial/ionization decomposition.","rationale":"I read the paper as a largely internally consistent numerical study of idealized turbulent boxes with non-equilibrium carbon chemistry. The strongest and most interesting claim is the complementarity: turbulence sets spatial structure, radiation sets ionization fractions. The load-bearing condition for that claim is that radiation is subdominant in determining where individual carbon species reside. Table 1 and Fig. 8 show this condition is violated for C IV: the radiation-induced change in the C IV density power-spectrum slope is nearly as large as the change produced by varying the turbulent velocity from 30 to 100 km/s. This is an internal tension, not an external modeling objection, so it is more directly decisive than the reader's concern about missing magnetic fields or compressive driving. The reader's conditional verdict is reasonable, but the condition should be sharpened: the manuscript should either weaken the central decomposition claim or quantify the relative contributions of U and sigma_3D to species spatial statistics. Because the simulations themselves are useful and the overstatement can be corrected by revision, I recommend conditional acceptance rather than rejection.","tokens_in":15195,"tokens_out":8956,"duration_ms":103760,"concrete_test":"Re-analyze Table 1 by computing, for each species, the radiation-induced slope change delta_R = |P_n(U=0) - P_n(U=1e-3)| at fixed sigma_3D and the turbulence-induced change delta_T = |P_n(sigma_3D=30) - P_n(sigma_3D=100)| at fixed U. If for C IV delta_R at sigma_3D=100 (1.66) is at least half of delta_T at U=0 (1.80), the central decomposition fails; report the analogous comparison for C II and C I. This requires no new simulations and directly tests the load-bearing claim using published numbers.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires a clean separation: turbulence sets where carbon species live, radiation sets only their ionization fractions. The paper's own Table 1 contradicts this for C IV, the principal observational tracer. At fixed sigma_3D = 100 km/s, adding radiation changes the C IV density power-spectrum slope from +0.87 (U=0) to -0.79 (U=1e-3), a shift of 1.66 in slope units. The entire turbulence-induced change from sigma_3D = 30 to 100 km/s at U=0 is only from -0.93 to +0.87, a shift of 1.80. Thus radiation is comparable to a factor-of-three change in turbulent velocity in setting the spatial clustering of C IV. The 2D PDF in Fig. 8 reinforces this: with radiation, n_CIV first rises and then falls with increasing density, a radiation-driven spatial bias caused by recombination in dense regions, not merely a change in the overall ionization fraction. Conclusion item 1, that turbulence is the primary driver of the spatial distribution of carbon species, is therefore not supported for C IV, and the proposed decomposition of observed C II/C IV absorption into a turbulence-dominated spatial term and a radiation-dominated ionization term overstates the separation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents three-dimensional hydrodynamic simulations of a turbulent, uniformly filled periodic box, using the MAIHEM non-equilibrium chemistry package, to study the spatial distribution and ionization balance of carbon species (C I, C II, C IV) in the circumgalactic medium. The simulations vary the turbulent velocity dispersion (σ_3D = 30, 60, and 100 km/s) and the ionizing background (U = 0 and U = 10^-3), and the authors analyze density and kinetic-energy power spectra, species density distributions, and correlations between density, species abundance, and viscous dissipation. The central claim is that turbulence is the primary driver of the spatial distribution of carbon species, while the background radiation mainly sets the relative ionization fractions, leading to a proposed decomposition of observed C II/C IV absorption into a turbulence-dominated spatial term and a radiation-dominated ionization term.","tokens_in":15505,"tokens_out":2514,"duration_ms":27778,"significance":"If the central claim holds, the paper offers a useful physically motivated framework for interpreting CGM absorption-line observations of carbon ions, and it extends earlier non-equilibrium chemistry work by the same group to a systematic study of how turbulent driving and ionizing background jointly set carbon structure. The use of a 65-ion non-equilibrium network and the comparison of solenoidal turbulence cases across three velocity dispersions are valuable, and the tabulated spectral slopes provide a compact summary of trends that can be compared with future simulations and observations. However, the significance is moderated by the idealized setup (uniform-density box, no magnetic fields, gravity, or outflows, as acknowledged in Section 3.5) and by the quantitative inconsistency between the paper's own Table 1 and the claimed separation of turbulence and radiation effects.","major_comments":[{"comment":"The claim in Conclusion item 1 that turbulence is the primary driver of the spatial distribution of carbon species is not supported for C IV by the paper's own spectral slope measurements. At fixed σ_3D = 100 km/s, adding radiation changes the C IV density power-spectrum slope from +0.87 (U = 0) to -0.79 (U = 10^-3), a shift of 1.66 in slope units. By comparison, changing σ_3D from 30 to 100 km/s at U = 0 shifts the C IV slope by only 1.80 units (-0.93 to +0.87). Thus radiation has an effect on the spatial clustering of C IV comparable to a factor-of-three change in turbulent velocity. The bimodal density dependence shown in Figure 8, where n_C IV first rises and then falls with increasing density when radiation is present, is a radiation-induced spatial bias, not merely a change in the overall ionization fraction. The proposed decomposition of observed C II/C IV absorption into a turbulence-dominated spatial term and a radiation-dominated ionization term therefore overstates the separation, at least for C IV.","section":"Table 1 and Section 3.3"},{"comment":"The spectral slopes in Table 1 and Figures 3, 6, and 7 are presented without uncertainties and without a stated fitting procedure or fitting range. Since the central quantitative comparisons (Kolmogorov versus Burgers scaling, effects of radiation on clustering) rest on these slopes, the absence of error bars or a defined fit interval makes it difficult to judge whether the reported differences, such as the total density slope changing from -1.02 to -0.50 across the three velocity dispersions, are numerically robust. I request a description of the fitting method, the wavenumber range used, and the associated uncertainties, as well as a convergence test at different grid resolutions.","section":"Section 3.1 and Table 1"},{"comment":"The simulations use a uniform-density, periodic box with purely solenoidal driving, no magnetic fields, no gravity, and no galactic outflows, and the limitations are acknowledged in Section 3.5. This is a legitimate modeling premise, but the paper does not provide any test of whether the conclusions are robust to the most basic deviations from it. In particular, because the results emphasize power-spectrum slopes and spatial clustering of trace ions, a resolution study and at least one run with a different driving geometry or with a compressive component would be needed to establish that the claimed turbulence-radiation decomposition is not an artifact of the chosen forcing and numerical dissipation.","section":"Section 2.1 and Section 3.5"}],"minor_comments":[{"comment":"The text refers to 'redshift zero HM2012 EUVB' but the reference for HM2012 is not included in the reference list; please add the citation (e.g., Haardt & Madau 2012).","section":"Section 2.1"},{"comment":"The definition of ΔX_i/X_i in Equation (2) is ambiguous: the superscripts a and b are not defined, and it is unclear whether the quantity is a fractional change per timestep or a convergence criterion evaluated over a fixed interval. Please clarify in the text.","section":"Equation (2)"},{"comment":"The caption contains the phrase 'taken our simulations' which appears to be a typo for 'taken from our simulations' or 'from our simulations.'","section":"Figure 8 caption"},{"comment":"The choice μ = -1 g cm^-1 s^-1 makes Φ a squared rate with units s^-2, but the text calls it 'energy loss per unit volume per unit time.' Since the magnitude of Φ is arbitrary and set by the numerical viscosity convention, correlations with Φ should be described as qualitative; the current wording may be misleading.","section":"Section 3.2 and Equations (6)-(8)"},{"comment":"The entries for Pan & Scannapieco 2010a and 2010b both list page 1765 and appear nearly identical; please verify that these are distinct papers and cite them accurately.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a straightforward extension of the authors' prior non-equilibrium chemistry simulation work, and the self-citation pattern is not inappropriate given the methodology. The main concern is not circularity but the quantitative tension between the paper's own Table 1 and its headline claim about turbulence-dominated spatial structure. I would support publication after the authors either substantially qualify the C IV spatial claim or reframe the central conclusion to match the presented data. The missing convergence study and error bars on spectral slopes should also be addressed before the paper can be considered fully convincing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi [name],\n\nI read the paper by Hu et al. on carbon in turbulent CGM. It's a straightforward simulation study, and if you work on CGM absorption lines it's worth your time. The new material is the species-resolved power spectra—CI, CII, CIV density and kinetic spectra across Mach number and with/without UV background. The result that CIV is collisionally rare and spatially locked to dense filaments without radiation, but becomes common and anti-correlated with density once a UV background is present, is a clean, useful addition. The overall density spectral slopes match the known isothermal turbulence results, which is a good check.\n\nThe soft spot is the headline interpretation. The abstract and Conclusion 1 say turbulence governs the spatial distribution while radiation only sets ionization fractions. But Table 1 shows radiation changes the CIV density power-spectrum slope from +0.87 to -0.79 at sigma_3D=100 km/s, a shift of 1.66. That's nearly as large as the effect of going from 30 to 100 km/s at zero radiation (1.80). So for CIV, radiation is not merely changing the overall abundance; it's actively moving where the ions sit, because recombination in dense gas wipes them out there. The paper actually shows this bimodality in Fig. 8, but the abstract/conclusion wording doesn't follow.\n\nAlso, the quoted slopes have no uncertainties and there's no resolution or convergence study. For a paper whose central contributions are spectral slopes, that is a real omission. The arbitrary viscosity normalization (mu = -1) is fine as a bookkeeping device, but it's worth saying more clearly that Phi is not the physical dissipation rate.\n\nThe setup is idealized—periodic box, solenoidal forcing, no fields, no outflows—and the authors acknowledge this in 3.5. That limits how directly the numbers transfer to real halos, but as a parameter-space exploration it's defensible.\n\nMy bottom line: the citations to the group's own chemistry code are legitimate, the results are emergent from the simulations, and the paper is worth a serious referee. I'd send it to review, but I'd push for a rewritten abstract that removes the strict spatial/ionization separation for CIV, and for some convergence evidence. If the authors fix that, this becomes a solid reference for mock absorption-line studies.","headline":"Useful species-resolved turbulence study, but the central spatial/ionization decomposition is overstated: radiation reshapes C IV clustering almost as much as turbulence does.","tokens_in":15990,"tokens_out":2839,"would_cite":true,"duration_ms":27483,"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":"Turbulence sets where carbon sits in the circumgalactic medium; radiation sets what charge it carries.","keywords":["circumgalactic medium","turbulence","non-equilibrium chemistry","carbon ionization","C II absorption","C IV absorption","Burgers turbulence","density power spectrum"],"falsifier":"In quasar absorption-line data, measure whether C IV absorption is less spatially clustered than C II at fixed column density, as the model predicts when a UV background is present; if C IV is found to be as clustered as C II in dense gas, the predicted recombination-driven anti-correlation at high density is wrong.","tokens_in":15031,"feed_emoji":"🌌","tokens_out":10360,"duration_ms":99268,"temperature":0.7,"pith_summary":"The paper sets out to explain why carbon absorption lines in the circumgalactic medium (CGM) look the way they do, and argues that two physical agents do two different jobs. Turbulence is the primary driver of the spatial distribution of carbon species: it compresses gas into filaments, creates shocks, and concentrates C II and C IV near regions of strong viscous dissipation, while the kinetic energy spectrum steepens from Kolmogorov to Burgers scaling as turbulent velocity rises. Background radiation, in contrast, sets the ionization balance: adding a UV background shifts the bulk of carbon from C I to C II and boosts C IV by roughly sixteen orders of magnitude, while pushing C IV into lower-density regions where recombination is slow. The upshot is that observed C II and C IV absorption can be decomposed into a turbulence-dominated spatial term and a radiation-dominated ionization term.","feed_headline":"Turbulence maps where carbon sits; radiation sets its charge","feed_subtitle":"Simulations show C II and C IV absorption can be read as separate spatial and ionization maps of the CGM.","key_machinery":"The central machinery is a suite of three-dimensional hydrodynamic simulations of a $512^3$ periodic box, $10\\,\\mathrm{kpc}$ on a side, filled with gas at uniform density and driven by solenoidal turbulence. The simulations evolve a 65-ion non-equilibrium chemistry network for hydrogen, helium, carbon, nitrogen, oxygen, neon, sodium, magnesium, silicon, sulfur, calcium, and iron, including photoionization by a UV background, and are run with and without the background to isolate its effect. The load-bearing diagnostics are power spectra: the kinetic energy spectrum $E_v(k)$ and the density spectrum $P_n(k)$ for the total gas and for C I, C II, and C IV separately, together with two-dimensional probability distributions of species density versus total density and versus the squared viscous dissipation rate $\\Phi = \\Phi_c + \\Phi_s$. Because the hydrodynamics and chemistry are invariant under the rescaling $x\\to\\lambda x$, $t\\to\\lambda t$, $\\rho\\to\\rho/\\lambda$, the results depend only on $nL$, $\\sigma_{1D}$, and the ionization parameter $U$, so the box can stand in for CGM gas of different density.","core_discovery":"On the paper's own terms, the discovery is a clean division of labor. In the turbulent CGM, turbulence controls where each carbon species is found: at velocity dispersions from 30 to 100 km/s, the gas becomes more filamentary, small-scale density fluctuations grow, and the total kinetic energy spectrum shifts from a Kolmogorov slope near $-5/3$ to a Burgers slope near $-2$, signaling shock-dominated dissipation. C II and C IV form preferentially in shock-heated, high-dissipation regions, and their density fields track the filamentary total-density field. Radiation then controls what fraction of carbon is in each ionization state: with an ionization parameter $U=10^{-3}$, most carbon becomes C II, C IV abundance rises enormously (about $10^{16}$ times relative to the no-radiation case), and C IV is biased toward low-density gas because recombination destroys it more rapidly in dense regions. The combination produces a bimodal density-C IV relation that is the paper's most distinctive signature of the turbulence-radiation interplay.","pith_inferences":["If the turbulence-radiation decomposition holds, the C II/C IV column-density ratio along a quasar sightline could be inverted to recover the local ionization parameter, while line-width differences between the two ions could recover the turbulent Mach number, effectively turning absorption spectra into two independent physical maps.","The same simulation machinery should apply to other ions in the 65-ion network, such as O VI or Si IV; if their density spectra show the same turbulence-versus-radiation split, the framework becomes a general tool for interpreting UV metal-line absorbers.","The bimodal density-C IV relation suggests an observational diagnostic: the density at which C IV switches from tracking to anti-tracking total density is a direct measure of the local UV background strength, so mapping that turnover across different halo environments could calibrate the metagalactic background.","Magnetic fields are the most likely disruptor: if small-scale turbulent dynamo action makes the density structures anisotropic, the isotropic spectral slopes reported here would need to be replaced by orientation-dependent ones, but the turbulence-radiation decomposition itself may survive."],"forward_implications":["At high turbulent velocity dispersion, the CGM density field becomes filamentary and its power spectrum flattens, so absorption-line components should appear more clustered on small scales as turbulence strengthens.","Without a UV background, C IV is essentially absent ($\\lesssim 10^{-20}\\,\\mathrm{cm^{-3}}$ even at 100 km/s), so any strong C IV detection implies either a substantial ionizing background or additional heating beyond solenoidal turbulence.","With a UV background, C IV traces low-density gas, so the ratio of C IV to C II column density is not a simple density tracer; it encodes the local competition between photoionization and recombination.","The kinetic energy spectrum's transition from Kolmogorov to Burgers scaling marks shock-dominated dissipation as a key heating mechanism, meaning turbulence itself can create ionized carbon even when radiation is weak.","The density correlation function steepens from roughly $\\xi_n\\propto r^{-2}$ toward $r^{-2.5}$ or steeper as turbulence rises, quantifying how much more clumpy the CGM becomes."],"supporting_citations":[{"why":"Supplies the non-equilibrium cooling and chemistry model for 65 ions that the simulations evolve.","marker":"Gray et al. (2015)"},{"why":"Derives the hydrodynamics-plus-chemistry equations and the scale-invariance transformation used to map results to CGM conditions.","marker":"Gray & Scannapieco (2016)"},{"why":"Extends the non-equilibrium ionization network and radiative-process treatment on which the carbon abundance calculations rely.","marker":"Gray & Scannapieco (2017)"},{"why":"Provides the detailed simulation methodology and setup that this paper follows.","marker":"Buie et al. (2018)"},{"why":"Establishes the solenoidal turbulent driving and the expected spectral behavior of isothermal turbulence used for comparison.","marker":"Pan & Scannapieco (2010b)"},{"why":"Offers the isothermal-turbulence spectral results used to check the kinetic-energy-spectrum slopes.","marker":"Federrath et al. (2010)"},{"why":"Gives the measured density power-spectrum slopes versus Mach number that anchor the interpretation of the slopes seen here.","marker":"Kim & Ryu (2005)"},{"why":"Provides the Burgers-turbulence model for density power spectra in strong- and weak-shock limits used to interpret the flattened density spectra.","marker":"Saichev & Woyczynski (1996)"}],"fun_headline_variants":["In CGM, turbulence places carbon; radiation sets its ionization","Turbulence vs radiation: how CGM carbon gets its spatial and ionized state","Shock-dominated turbulence and background radiation jointly shape carbon ions","Two levers for CGM carbon: turbulent flow and ionizing light","Simulations reveal dual control of carbon phases in turbulent CGM"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results assume the CGM behaves as a uniform gas cube in which turbulence is driven only by vortex motion, with no magnetic fields, gravity, or galactic outflows; if real CGM turbulence is driven by compression or affected by magnetic fields, the predicted patterns may not hold in actual halos.","fun_headline_variants_meta":{"raw":{"variants":["In CGM, turbulence places carbon; radiation sets its ionization","Turbulence vs radiation: how CGM carbon gets its spatial and ionized state","Shock-dominated turbulence and background radiation jointly shape carbon ions","Two levers for CGM carbon: turbulent flow and ionizing light","Simulations reveal dual control of carbon phases in turbulent CGM"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000871,"raw_usage":{"total_tokens":3772,"prompt_tokens":947,"completion_tokens":2825,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":563,"completion_tokens_details":{"reasoning_tokens":2733}},"tokens_in":563,"tokens_out":2825,"duration_ms":22280,"temperature":1.0,"reasoning_tokens":2733,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:07:30.847913+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In quasar absorption-line data, measure whether C IV absorption is less spatially clustered than C II at fixed column density, as the model predicts when a UV background is present; if C IV is found to be as clustered as C II in dense gas, the predicted recombination-driven anti-correlation at high density is wrong.","supporting_citations":[{"cited_title":"J., Scannapieco , E., & Kasen , D","cited_arxiv_id":null,"evidence_quote":"Supplies the non-equilibrium cooling and chemistry model for 65 ions that the simulations evolve."},{"cited_title":"J., & Scannapieco , E","cited_arxiv_id":null,"evidence_quote":"Derives the hydrodynamics-plus-chemistry equations and the scale-invariance transformation used to map results to CGM conditions."},{"cited_title":"I., & Woyczynski, W","cited_arxiv_id":null,"evidence_quote":"Provides the Burgers-turbulence model for density power spectra in strong- and weak-shock limits used to interpret the flattened density spectra."}],"review_version":2}