{"id":"d97e60e2-1a28-4a3d-aeed-38e9d50c1ee8","arxiv_id":"2504.17842","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Magnetic fields in the western Eta Carinae giant molecular cloud are strong enough to keep most massive clumps subcritical, while a column-density threshold marks where gravity takes over.","lead":"Using infrared polarization data from the SOFIA telescope, this paper maps magnetic fields across 17 massive star-forming clumps in the western Eta Carinae cloud. The fields appear strong enough to hold most of the gas against gravity, which helps explain why these clumps form stars slowly.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (3)'s IRG depth estimate is uncalibrated and caps at the beam near column-density peaks, so DCF-based mass-to-flux ratios are biased low exactly where gravity should dominate; the HRO also never clearly reaches ξ<0, leaving the subcritical claim dependent on this biased λ.","rationale":"I agree with the reader that the IRG depth estimate is the weakest load-bearing step. It is not a nuisance parameter: it enters every DCF field strength and mass-to-flux value, and the cap at the beam size applies at column-density peaks where star formation actually proceeds. The direction of the bias is unfavorable: R is likely underestimated at peaks, so λ is biased low, potentially hiding supercritical regions. This cannot be dismissed by the HRO because, as the paper itself states, the full Region 9 HRO never reaches a clearly perpendicular regime; the quoted Ncrit is the zero crossing of a fitted line, not a regime of observed ξ < 0. The paper does include several internal checks—the 0.5 pc correlation-length variant, the Skalidis & Tassis (2021) formula, and the consistency of logλ trends with Tdust—and these make the qualitative picture plausible. However, all these checks inherit the same R map when they use Eqs. (1) and (2), so they do not test the depth assumption. The proposed recomputation with an independent depth scale would settle whether the central claim survives. Because the reader's conditional verdict already requires exactly this kind of test, I do not change the verdict.","tokens_in":23930,"tokens_out":9125,"duration_ms":100704,"concrete_test":"Recompute the DCF B⊥ and logλ maps using an independent, clump-by-clump depth estimate, e.g., R = deconvolved geometric mean of the 12CO clump major/minor axes from Barnes et al. (2016) or Pitts et al. (2019), or an excitation-temperature/opacity-based 13CO column-to-volume conversion, instead of the IRG of Eq. (3). Compare the resulting mean logλ, the fraction of pixels with logλ > 0, and the binned B–n trend. If the mean logλ rises by more than ~0.3 dex, or the supercritical-pixel fraction exceeds a few times 10%, the conclusion that magnetic fields keep Region 9 subcritical is not robust to the depth assumption. As a minimal check, propagate a systematic factor of 2–4 increase in R at all pixels with negative Laplacian and report how the headline statistics shift.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central quantitative claim—that magnetic fields provide enough support to explain the low star formation rate—rests on the DCF-derived mass-to-flux map. Every B⊥ and λ value inherits the density n = N_H2/R (Eq. 3), where the line-of-sight depth R is estimated from the inverse relative gradient (IRG), with R set to the 0.16 pc beam wherever the column-density Laplacian is negative. This depth is never calibrated against an independent measure of cloud depth. Near column-density peaks, exactly the pixels most relevant to star formation, the true line-of-sight depth of a clump is generally comparable to its diameter (often >1 pc), not the beam size. Because λ ∝ sqrt(R) (and B⊥ ∝ 1/sqrt(R)) at fixed N, capping R at 0.16 pc when the true R is ~1 pc suppresses λ by about a factor of 2.5 (0.4 dex), and by more for larger clumps. Such a bias acts precisely where gravity is expected to dominate, increasing the supercritical fraction and shifting the mean logλ from –0.75 toward values closer to zero or above. The absolute B values in Fig. 16 and the comparison to the Crutcher (2012) relation are also R-dependent, so the inference of 'somewhat higher than typical' field strengths is not robust. The HRO analysis is independent of R, but as the text itself admits in §3.2, the Region 9 data do not clearly reach ξ < 0; the quoted crossing at log N = 26.56 is an extrapolation of a negative slope from mostly-parallel bins. Thus the HRO cannot independently rescue the quantitative subcritical claim if the IRG depth assumption is wrong.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents SOFIA/HAWC+ 154 μm polarimetry of 17 massive clumps in Region 9 of the CHaMP survey at the western end of the η Carinae GMC, observed at a fixed physical resolution of 0.16 pc. The authors apply Davis-Chandrasekhar-Fermi (DCF) and Histogram of Relative Orientation (HRO) analyses to map the plane-of-sky magnetic field strength B⊥, the mass-to-flux ratio λ, and the field alignment relative to column-density structures. They report a mostly subcritical cloud (mean log λ = -0.75 ± 0.45 under classical DCF), an HRO transition at log N_crit = 26.56 ± 0.07, and B–n values lying somewhat above the Crutcher (2012) relation. They conclude that magnetic fields provide enough support against gravity to explain the low star formation rate in these massive clumps, except in regions strongly affected by feedback from the HII region NGC 3324.","tokens_in":24299,"tokens_out":8062,"duration_ms":73246,"significance":"If the central quantitative result survives scrutiny, this is a valuable contribution to massive star formation studies: it provides one of the first systematic, fixed-resolution surveys of magnetic support across a single GMC containing clumps in all evolutionary stages. The dataset is substantial (about 9000 independent polarization measurements), the analysis uses two standard and partially independent methods that both indicate the importance of magnetic support, and the authors are commendably explicit about key limitations, including the built-in B–n correlation in DCF and the absence of a direct ξ<0 detection in the HRO data. The HRO transition column is a quantitative, falsifiable prediction that can be tested with Zeeman or higher-resolution polarimetry. However, the quantitative subcritical conclusion rests on an uncalibrated line-of-sight depth estimate that needs a sensitivity analysis before the results can be fully trusted.","major_comments":[{"comment":"The IRG-based depth estimate is the load-bearing input for all quantitative DCF results, but it is never calibrated against an independent measure of cloud depth. Because B⊥ ∝ sqrt(N/R) (Eq. 1) and λ ∝ N/B_TOT ∝ sqrt(N R) (Eq. 4), capping R at the 0.16 pc beam wherever the Laplacian is negative suppresses λ near column-density peaks by roughly sqrt(0.16/R_true), i.e., about 0.4 dex for a typical clump depth of 1 pc, and more for larger clumps. This bias acts exactly in the regions where gravity is expected to dominate, so the reported mean log λ = -0.75 and the statement that 'only small areas are dominated by gravity' may be substantially in error. I recommend (a) validating the IRG-derived R against independent depth estimates for at least a subset of clumps (e.g., sizes from the Mopra 12CO ellipses, or column-density/volume-density comparisons), and (b) recomputing the λ maps and the supercritical fraction under a range of plausible R assumptions (e.g., constant R per clump, R = projected diameter, or a factor 2–3 larger near peaks) to demonstrate the robustness of the subcritical conclusion.","section":"§3.1, Eq. (3)"},{"comment":"The quoted threshold log N_crit = 26.56 ± 0.07 is the zero-crossing of a linear fit to the HRO shape parameter ξ(N), but the text explicitly states that the data 'do not clearly progress to a column density regime where the alignments are more perpendicular (i.e., ξ < 0)'. The crossing point is therefore an extrapolation beyond the range of directly sampled ξ values (the highest bins have ξ consistent with zero, not negative). The abstract's claim that the threshold 'indicat[es] that gravitational forces exceed magnetic forces above this value' overstates the direct evidence. Please rephrase the claim to reflect that the data show a strong trend toward criticality at this column density, with the actual crossing being an extrapolation, and consider reporting the column density at which ξ becomes consistent with zero at the 1σ level rather than the formal zero-crossing.","section":"§3.2 and Abstract"},{"comment":"The B–n diagram analysis compares the fitted slope κ ≈ 0.5 to the Crutcher (2012) relation, but this slope is largely built into the construction: from Eq. (1) with n = N/R, B⊥ ∝ sqrt(n) at fixed s and ΔV, so the expected κ is 0.5 regardless of the actual physics. The authors correctly acknowledge this via Pattle et al. (2022) and present the Alfvén number as an alternative, but they still use the vertical offset ('a factor of ~2.5 above the Crutcher line') to infer 'somewhat higher than typical' field strengths. This offset is the only physically meaningful signal, and it depends directly on the assumed values of Q, L_corr, and R. Please provide a quantitative decomposition: how much of the offset is due to the chosen Q=0.5 and B_TOT=2B⊥ assumptions, and how much survives if R is varied as in the sensitivity analysis above? Without this, the comparison to Crutcher (2012) is difficult to interpret.","section":"§3.3, Figure 16 and surrounding text"}],"minor_comments":[{"comment":"The chain of equalities in Eq. (3) is visually confusing because n is written both as N_H2/R and as ∇N_H2; please rewrite to show the two branches separately (one for ∇²N_H2 > 0, one for ∇²N_H2 ≤ 0).","section":"Eq. (3)"},{"comment":"The text refers to 'the other 7 ROIs' as a single group, but BYF70aN is subsequently described as a separate outlier; please clarify the grouping, e.g., '6 flat ROIs plus 1 outlier, BYF70aN'.","section":"§3.3"},{"comment":"The figure label reports χ² = 0.59 for the binned fit; please specify whether this is the reduced chi-squared and state the number of degrees of freedom.","section":"Figure 16"},{"comment":"The t-statistics use sample sizes divided by 2.5² to account for oversampling, but the λ values are spatially correlated on the DCF correlation scale; a more conservative estimate using the number of independent resolution elements would strengthen the significance claim.","section":"Table 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid observational contribution, and the referee concerns are addressable with additional analysis rather than new observations. I would like the revision to include a sensitivity analysis for the depth estimate in Eq. (3), a softened interpretation of the HRO zero-crossing, and a decomposition of the B–n offset. The dataset itself is valuable and fits the journal's scope well."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Mark,\n\nQuick take: this is a genuinely useful observing paper—new SOFIA/HAWC+ polarization data for 17 massive clumps in the western Eta Car GMC, all at the same 2.5 kpc distance, analyzed with DCF and HRO. The two methods point the same way: field alignments change from parallel to perpendicular with column density, and the mass-to-flux ratio is mostly subcritical. That consistency, plus the transparent acknowledgment of known correlations, makes the qualitative story believable.\n\nWhat's actually new: 15 of 17 clumps have no previous polarization maps. The nominal vs. flat HRO split, with the flat group plausibly shaped by NGC 3324, is a nice environmental result. The Tdust-lambda trend is also a new observational correlation, and it ties to the earlier Pitts et al. work. The paper is well organized and the figures are informative.\n\nThe soft spots are in the quantitative layer. The DCF analysis converts column to volume density using the inverse relative gradient (IRG) as depth, and caps R at the 0.16 pc beam wherever the Laplacian is negative. That cap sits exactly on column density peaks, where the true line-of-sight depth is probably closer to the clump diameter than to a beam. Since lambda ∝ sqrt(R) at fixed N, the cap suppresses lambda by a factor of about 2.5 (0.4 dex) where gravity should be strongest. That doesn't necessarily kill the subcritical conclusion—the mean would move from -0.75 to maybe -0.3 or -0.4—but it does undermine the specific claim that B fields alone explain the low SFR. The same issue biases the B-n comparison and the 'somewhat higher than typical' field strengths. The authors should test sensitivity to a different depth prescription (e.g., uniform ellipsoid depths) or calibrate against an independent measure, if any exists.\n\nThe HRO analysis is independent of R, but it never clearly reaches ξ < 0. The quoted Ncrit = 26.56 is a linear extrapolation of a negative slope from mostly parallel bins. The paper actually says this in §3.2, which is honest, but it means the HRO cannot independently rescue the quantitative criticality threshold.\n\nThe B-n built-in correlation is acknowledged, and the Skalidis-Tassis comparison helps. The hand-picked correlation length (0.33 pc) and Q = 0.5 are standard choices; not a flaw, but they add uncertainty to absolute B values.\n\nBottom line: this deserves a serious referee. The new data are valuable, the qualitative picture is likely right, and the authors are upfront about the method's limitations. The referee should push on the depth estimate and on what can actually be concluded about the transition density. For a reader working on massive star formation and magnetic fields, this is a useful paper to have in the conversation. I'd accept for review, with major revisions expected.","headline":"New HAWC+ maps give the field a homogeneous 17-clump sample at one distance, and the subcritical trend is probably real; but the DCF-based mass-to-flux numbers rest on a depth estimate that is uncalibrated and biased exactly where it matters.","tokens_in":24944,"tokens_out":3170,"would_cite":true,"duration_ms":31565,"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":"This paper argues that magnetic fields provide enough support against gravity to explain the low star formation rate of the massive clumps in the western Eta Carinae molecular cloud, and identifies the column density threshold where…","keywords":["magnetic fields","star formation","molecular clouds","dust polarization","Davis-Chandrasekhar-Fermi analysis","histogram of relative orientations","mass-to-flux ratio","HII region feedback"],"falsifier":"Pick one clump, say BYF68, and obtain an independent line-of-sight depth (for instance, from molecular-line excitation ratios or from the sizes of its embedded cores), then recompute the DCF $B_\\perp$ and $\\lambda$ maps with those depths instead of the inverse-relative-gradient values. If the resulting mean $\\log\\lambda$ shifts by more than about 0.3 or moves across the subcritical-to-supercritical boundary, the paper's conclusion is not stable under its depth assumption; if it does not, the assumption is validated for that clump.","tokens_in":23714,"feed_emoji":"🧲","tokens_out":12040,"duration_ms":99999,"temperature":0.7,"pith_summary":"This paper uses far-infrared polarization maps of 17 massive molecular clumps at the western end of the Eta Carinae giant molecular cloud to ask whether magnetic fields can account for the region's low star formation rate. It finds a clear threshold: the field runs mostly parallel to dense gas structures below $\\log N_{\\rm crit} = 26.56 \\pm 0.07$, and mostly perpendicular above it, marking the column density where gravity overtakes magnetic support. Combining Davis\\textendash Chandrasekhar\\textendash Fermi field-strength estimates with Herschel column densities, most of the mapped area is subcritical, with a mass-to-flux ratio of $\\log \\lambda = -0.75 \\pm 0.45$; only about 2% of pixels exceed $\\lambda = 3$. Ten of the seventeen clumps follow the standard alignment trend, while seven show flat or reversed trends that the authors attribute to the nearby HII region NGC 3324. The paper's conclusion is that magnetic support, not a lack of dense gas, is what keeps these massive clumps from forming stars faster.","feed_headline":"Magnetic fields hold back star formation in the Eta Carinae cloud","feed_subtitle":"New far-infrared polarization maps show most gas stays magnetically supported; gravity wins only in cold cores.","key_machinery":"The argument is carried by two statistical tools and one geometric estimate. The HRO shape parameter $\\xi$ compares the orientation of the magnetic field to the tangent of column-density contours, with $\\xi>0$ meaning parallel and $\\xi<0$ perpendicular; a linear fit $\\xi = C_{\\rm HRO}(\\log N - X_{\\rm HRO})$ locates the critical column density where the field alignment switches. The DCF method converts the dispersion of polarization angles into a plane-of-sky field strength $B_\\perp$, and the mass-to-flux ratio $\\lambda = 6.38\\,(N_{\\rm H_2}/10^{26}\\,{\\rm m^{-2}})/(B_{\\rm TOT}/{\\rm nT})$ measures whether gravity or magnetism dominates. The geometric estimate is the 'inverse relative gradient' depth $R = N_{\\rm H_2}/\\nabla N_{\\rm H_2}$, used to convert column density into volume density; the paper itself notes this fails near column-density peaks, where it substitutes the projected beam size when the Laplacian is negative. This $R$ value feeds every $B_\\perp$ and $\\lambda$ map, so it is the main structural assumption in the analysis.","core_discovery":"The central claim is that in this sample of parsec-scale massive clumps, magnetic fields are strong enough to hold gravity in check over most of the mapped area, which is why star formation proceeds slowly despite large reservoirs of dense gas. The evidence has two independent legs. Histogram of relative orientation (HRO) analysis shows the expected transition from parallel to perpendicular field alignment with increasing column density, at $N_{\\rm crit} = (3.7\\pm0.6)\\times10^{26}\\,{\\rm m^{-2}}$ for the whole region and at a similar threshold in the ten 'nominal' clumps; the other seven clumps are exceptions whose fields are controlled by the radiation and overpressure of NGC 3324. Davis\\textendash Chandrasekhar\\textendash Fermi (DCF) analysis gives field strengths of roughly 10\\textendash 200 nT, with a peak of 469 nT, placing the $B$\\textendash$n$ data mostly above the Crutcher (2012) relation and giving a mostly subcritical mass-to-flux ratio. The paper also shows that colder dust is statistically much more likely to sit in supercritical gas, linking the magnetic criticality threshold to the known trend of dust temperature falling toward clump centers.","pith_inferences":["A testable extension is to use the HRO slope split (nominal vs. flat or reversed) as a diagnostic of external feedback in other clouds near HII regions; the sample here is small, but the pattern is specific enough to look for elsewhere.","Because $B_\\perp \\propto n^{1/2}$, a systematic error in the inverse-relative-gradient depth shifts inferred field strengths only weakly, but it shifts $\\lambda$ directly; an independent depth calibration for even one clump would show whether the subcritical mean is real.","The $\\log\\lambda$\\textendash$T_{\\rm dust}$ correlation suggests a prediction: at higher angular resolution, the gas that is already supercritical should coincide with the coldest, densest pixels, and the transition should sharpen toward clump centers; this can be checked with ALMA or next-generation far-infrared polarimetry.","Implicitly, if the critical column density shifts with environment, star-formation efficiency in massive clumps may be regulated by the local radiation field as much as by the magnetic field\\textendash gravity balance; comparing $N_{\\rm crit}$ in isolated massive clumps with this feedback-affected region would clarify that."],"forward_implications":["If magnetic support is the main brake on star formation, then the low star formation efficiency of the region is not caused by a shortage of dense gas but by the field's resistance to collapse; star formation should be concentrated in the few cold spots where $\\lambda$ exceeds 1.","The high critical column density ($\\log N_{\\rm crit} = 26.56$) compared with nearer, lower-mass clouds implies that massive clumps can remain magnetically supported up to higher column densities, pushing the onset of collapse to denser, colder gas.","The split between ten 'nominal' clumps and seven flat or reversed clumps indicates that external feedback from an HII region can locally reset the magnetic field\\textendash gravity balance, so environment, not just column density, determines where stars form.","The temperature trend, with the fraction of pixels having $\\log\\lambda>0$ rising from 0.08% in the warmest quintile to 12.4% in the coolest, ties the magnetic criticality threshold to dust temperature, consistent with the known $T_{\\rm dust}$\\textendash$N_{\\rm H_2}$ anticorrelation toward clump centers.","Strong fields ($B>100$ nT) appear in 5% of the high-signal pixels in just a fraction of one GMC, so the 'high-field' part of the $B$\\textendash$n$ diagram may be more populated than previously thought, with correspondingly higher critical field strengths for massive star formation."],"supporting_citations":[{"why":"Establishes the DCF relation that turns polarization-angle dispersion into a field-strength estimate, the backbone of the $B_\\perp$ maps.","marker":"Chandrasekhar & Fermi (1953)"},{"why":"Supplies the Zeeman-based $B$\\textendash$n$ relation and the criticality framework the paper compares its $B$\\textendash$n$ data against.","marker":"Crutcher (2012)"},{"why":"Defines the HRO shape parameter $\\xi$ and the regression form used to identify the critical column density.","marker":"Soler et al. (2017)"},{"why":"Provides the HRO methodology and its earlier application to BLASTpol data that this paper extends.","marker":"Fissel et al. (2016)"},{"why":"Provides the Herschel-derived $N_{\\rm H_2}$ and $T_{\\rm dust}$ maps that are the structural inputs for both analyses.","marker":"Pitts et al. (2019)"},{"why":"Prior HAWC+ and ALMA study of BYF73 whose DCF and HRO pipeline is applied here to the full clump sample.","marker":"Barnes et al. (2023)"},{"why":"Offers the compressible-MHD DCF variant used as a cross-check on $B_\\perp$ and $\\lambda$.","marker":"Skalidis & Tassis (2021)"},{"why":"Reviews DCF caveats and the built-in $\\kappa=1/2$ correlation in $B$\\textendash$n$ diagrams, used to interpret the high-density turnover.","marker":"Pattle et al. (2022)"},{"why":"Provides the large-scale parallel-to-perpendicular HRO transition at low resolution that the sub-parsec results are compared with.","marker":"Planck Collaboration XXXV (2016)"}],"fun_headline_variants":["Magnetic fields stall star formation in Eta Carinae clumps","Eta Carinae's magnetic fields thwart gravity, slow star birth","Most Eta Carinae gas magnetically braced; only cold cores collapse","Gravity loses to magnetism across Eta Carinae's massive clumps","Magnetic support explains slow star formation in Eta Carinae"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the line-of-sight depth of the gas, estimated as the inverse relative gradient of the Herschel column-density map ($R = N_{\\rm H_2}/\\nabla N_{\\rm H_2}$, with a beam-size substitution at column-density peaks), is a fair approximation to the true cloud depth; every $B_\\perp$ and $\\lambda$ map inherits this assumption, and the paper does not calibrate it against any independent depth measurement.","fun_headline_variants_meta":{"raw":{"variants":["Magnetic fields stall star formation in Eta Carinae clumps","Eta Carinae's magnetic fields thwart gravity, slow star birth","Most Eta Carinae gas magnetically braced; only cold cores collapse","Gravity loses to magnetism across Eta Carinae's massive clumps","Magnetic support explains slow star formation in Eta Carinae"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000525,"raw_usage":{"total_tokens":2687,"prompt_tokens":1246,"completion_tokens":1441,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":862,"completion_tokens_details":{"reasoning_tokens":1348}},"tokens_in":862,"tokens_out":1441,"duration_ms":10253,"temperature":1.0,"reasoning_tokens":1348,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:33:28.032681+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Pick one clump, say BYF68, and obtain an independent line-of-sight depth (for instance, from molecular-line excitation ratios or from the sizes of its embedded cores), then recompute the DCF $B_\\perp$ and $\\lambda$ maps with those depths instead of the inverse-relative-gradient values. If the resulting mean $\\log\\lambda$ shifts by more than about 0.3 or moves across the subcritical-to-supercritical boundary, the paper's conclusion is not stable under its depth assumption; if it does not, the assumption is validated for that clump.","supporting_citations":[{"cited_title":"1953, , 118, 113","cited_arxiv_id":null,"evidence_quote":"Establishes the DCF relation that turns polarization-angle dispersion into a field-strength estimate, the backbone of the $B_\\perp$ maps."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Zeeman-based $B$\\textendash$n$ relation and the criticality framework the paper compares its $B$\\textendash$n$ data against."},{"cited_title":"2017, A&A, 603, A64","cited_arxiv_id":null,"evidence_quote":"Defines the HRO shape parameter $\\xi$ and the regression form used to identify the critical column density."},{"cited_title":"M., Ade, P","cited_arxiv_id":null,"evidence_quote":"Provides the HRO methodology and its earlier application to BLASTpol data that this paper extends."},{"cited_title":"L., Barnes, P","cited_arxiv_id":null,"evidence_quote":"Provides the Herschel-derived $N_{\\rm H_2}$ and $T_{\\rm dust}$ maps that are the structural inputs for both analyses."},{"cited_title":"J., Ryder, S","cited_arxiv_id":null,"evidence_quote":"Prior HAWC+ and ALMA study of BYF73 whose DCF and HRO pipeline is applied here to the full clump sample."},{"cited_title":"& Tassis, K., 2021, A&A, 647, A186","cited_arxiv_id":null,"evidence_quote":"Offers the compressible-MHD DCF variant used as a cross-check on $B_\\perp$ and $\\lambda$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the large-scale parallel-to-perpendicular HRO transition at low resolution that the sub-parsec results are compared with."}],"review_version":1}