{"id":"f672e403-2e03-4520-a3ba-8bac6fa9d744","arxiv_id":"2502.08956","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A 20-band, 368,524-star Gaia-based analysis of the Coalsack molecular cloud finds optical and near-infrared extinction consistent with R_V=3.1, mid-infrared extinction matching R_V=5.5, and little variation of R_V with reddening above E(B-V)=0.3 mag.","lead":"This paper measures how dust reddens starlight across 20 optical and infrared bands toward the Coalsack, a nearby quiet, starless molecular cloud, using hundreds of thousands of Gaia-selected dwarf stars. It finds that the cloud follows the standard Milky Way extinction law (R_V about 3.1) in optical and near-infrared light and a flatter, dense-cloud law (R_V about 5.5) in the mid-infrared.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"MIR 'R_V = 5.5' conclusion rests on sparse, high-extinction MIR samples; robustness check needed.","rationale":"The reader's weakest_assumption focused on the C0 intrinsic-color calibration, which is genuinely load-bearing. My concern is more specific: the MIR R_V=5.5 claim is the least secure sub-claim because it depends on the smallest sample sizes (W3: 229 sources; dense GLIMPSE: ~100 sources) and on the same C0 calibration but with greater sensitivity to systematic offsets. The paper's internal evidence (Table 2, Section 3.2) shows sparse high-extinction points are retained without a minimum-count requirement, and the quoted errors exclude C0 systematics. A direct robustness test—recomputing the MIR CERs with stricter bin requirements and propagated calibration errors—would settle whether the MIR conclusion is solid or model-dependent. I agree with the reader's CONDITIONAL verdict but identify a more concrete mechanism within the MIR regime, hence 'partial' agreement.","tokens_in":19145,"tokens_out":1513,"duration_ms":14670,"concrete_test":"Recompute the MIR CERs and A_lambda/A_KS after (a) requiring at least 3 sources per 0.01-mag bin in the high-extinction tail, (b) propagating the intrinsic-color zero-point uncertainty into the CER errors, and (c) comparing the W3-based A_W3/A_KS against the GLIMPSE [3.6]–[8.0] values. If the W3 CER shifts by more than ~1 sigma when high-extinction sources are clipped, or if the MIR relative extinction moves away from the WD01 R_V=5.5 curve, the MIR claim weakens.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that the Coalsack extinction law follows R_V=5.5 in the MIR rests on converting measured CERs (e.g., E_J,W3/E_J,KS) into relative extinction using the WC19/WD01 extinction-law family. The weakest link is the very small number of high-extinction sources in the MIR bands, especially W3 and GLIMPSE. Table 2 shows the W3 Coalsack sample is only 229 sources total; the dense-region samples for GLIMPSE bands are ~90–150 sources. Section 3.2 states that high-extinction bins (E_J,KS > 0.5 mag) keep all points regardless of count, so the fit can be dominated by a handful of outliers. The quoted CER errors are from the linear fit and do not include the systematic uncertainty in the intrinsic-color calibration C0, which is derived from the bluest 3% of stars with AG<0.05 mag. Any residual reddening or Teff-dependent bias in the blue-edge C0—which the authors themselves note diverges below 4500 K and above 7000 K—propagates directly into the MIR CERs. The u-band deviation (Section 4.1.1) shows that the blue-edge method can produce wavelength-dependent systematics, and the MIR R_V=5.5 attribution is more sensitive to such shifts than the optical R_V=3.1 result because the MIR CER ratios are close to the WD01 curve and the sample is much sparser. Thus the MIR conclusion is the least secure component of the central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a multi-wavelength (0.35–12 μm) extinction law for the Coalsack molecular cloud, a quiescent starless cloud, using 368,524 Gaia DR3 dwarf stars as tracers and photometry from APASS, SMSS, 2MASS, GLIMPSE and WISE across 20 bands. The authors derive Teff–intrinsic-color relations via the blue-edge method, compute color excesses in each band, fit CE–CE diagrams to obtain color excess ratios (CERs), and convert these to relative extinction Aλ/AV and Aλ/AKS using the Wang & Chen (2019) extinction law. The main results are: (i) the optical–NIR extinction law follows R_V=3.1; (ii) the MIR law is flat and follows WD01 R_V=5.5; (iii) the E(B–V) maps at 1.3' resolution show fine structure and agree broadly with Guo et al. (2022); (iv) there is no strong correlation between R_V and E(B–V) for E(B–V)>0.3 mag, with a mean R_V=3.24±0.32.","tokens_in":19380,"tokens_out":15761,"duration_ms":140187,"significance":"If the results are robust, this is the first comprehensive optical–MIR extinction-law study of a quiescent starless cloud, and the finding that the Coalsack matches R_V=3.1 in the optical–NIR and R_V=5.5 in the MIR, like active star-forming clouds, is an interesting environmental comparison. The 1.3'-resolution extinction map and the multi-band CER catalogue are useful products for the community. The paper includes careful consistency checks: the E(B–V) map agrees with Guo et al. (2022) to ~0.03 mag, and the A_V comparison with Dobashi et al. (2005) is discussed with a plausible explanation. The main caveats are the calibration dependence of the R_V conversion and the small high-extinction samples, which require additional sensitivity analysis.","major_comments":[{"comment":"The Teff–C0 relations are built from the bluest 3% of stars in the ICR with AG<0.05 mag, but the ICR (296°≤l≤312°, –5°≤b≤15°) contains the Coalsack cloud itself, so residual reddening in the blue edge cannot be excluded. The uncertainties in Table 2 are only the linear-fit errors and do not include systematic errors from the C0 calibration; the polynomial fits are stated to diverge outside 4500–7000 K, and the u-band deviation discussed in §4.1.1 shows that the blue-edge method produces wavelength-dependent systematics. Please quantify how the CERs (and thus the R_V=3.1 and R_V=5.5 conclusions) shift when C0 is re-derived with, e.g., the bluest 1% or 5% thresholds, or when the Teff scale is shifted by the GSP-Phot median error of 119 K, and add a systematic term to the reported errors.","section":"§3.1, Figures 2–3"},{"comment":"Equations (2)–(3) convert the measured CERs into Aλ/AV and Aλ/AKS using AGBP/AGRP from WC19, and §4.4 maps CERs to R_V using the WC19/WC23 R_V-dependent family. Consequently, the agreement with R_V=3.1 in the optical-NIR and with R_V=5.5 in the MIR is not an independent test of R_V; it tests consistency with the same model family used for the calibration. The sensitivity is large: for the Coalsack values k_J=0.769 and k_KS=1.165, changing AGBP/AGRP from 1.7 to 1.5 changes AJ/AKS from about 2.7 to about 1.5, which propagates into all Aλ/AKS entries in Table 3 and shifts the inferred MIR R_V. Please report the derived extinction law and R_V for at least one alternative choice of AGBP/AGRP (e.g., HD20 or a WD01 R_V=5.5 curve) and state explicitly that the R_V values are conditional on the adopted calibration.","section":"§3.3, §4.4, Eqs. (2)–(3)"},{"comment":"The inner dense region (EGBP,GRP≥1.25 or EJ,KS≥0.5) has very small samples: 10 sources for V, 81–149 for the GLIMPSE and WISE bands, and the W3 Coalsack sample is only 229 sources in total. The fitting procedure keeps all high-extinction points regardless of source count, so the dense-region CERs (e.g., V: –1.050±0.080; H: 0.566±0.050) could be driven by a few outliers. These CERs are the basis for the claimed NIR-MIR regional variation in Fig. 7 and for the weak R_V–EB,V trend in Fig. 11. Please give per-bin source counts, run a robust fit (e.g., median-based or Theil-Sen), and show the dense-region CERs with and without the highest-extinction points; also verify the W3 point with a binned fit.","section":"§3.2, Table 2"},{"comment":"There is a direct inconsistency in the quoted Coalsack sample size: Section 2.3 states 'the final Coalsack, Ref. 1 and Ref. 2 samples contain 4,757, 117,585 and 32,964 stars', but Table 1 lists Coalsack as 14,112 (and Ref. 2 as 32,964). In addition, the Coalsack boundary in the text (299°≤l≤306°, –4°≤b≤2°) differs from Table 1 (l=299°~305.5°, b=–2.8°~2.45°). Please correct the text, make the boundary definition consistent, and confirm which value (4,757 or 14,112) was used in the analysis.","section":"§2.3, Table 1"}],"minor_comments":[{"comment":"The notation for color excess is inconsistent: 'EB,V' appears in the abstract while 'E(B–V)' or \"E_{B,V}\" is used elsewhere; please unify to a single subscripted form.","section":"Abstract and throughout"},{"comment":"The caption reads 'Multi-wavelengt Color Excess Ratios'; the typo 'wavelengt' should be corrected to 'wavelength'.","section":"Table 2 caption"},{"comment":"The header 'Aλ/Av' should be 'Aλ/AV' (italic V subscript) to match the text.","section":"Table 3 header"},{"comment":"There is a full-width comma in 'In contrast ，the ADobashi+05 V,mean values'; please use a standard comma.","section":"§4.3"},{"comment":"The reported 'R_V=3.24±0.32' is the mean and dispersion of the sub-region median R_V values, not the uncertainty of the mean; please state this explicitly and consider weighting by the number of stars per sub-region.","section":"§4.4, Figure 13"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid observational study with a clear methodology and useful products, but the central R_V claims depend on the adopted calibration (WC19/WC23) and on the uncertain blue-edge intrinsic-color relations. The sample-size inconsistency in §2.3 vs Table 1 must be resolved. I have not questioned the authors' expertise or the field's standard use of WC19; the request is for sensitivity analyses, not for a different method. The manuscript is within the journal's scope and should be reconsidered after major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things worth knowing before you read this. The real product is the measured color-excess ratios and the 1.3-arcmin extinction map for a cloud that lacked a full multi-wavelength characterization. The headline R_V values are partial circularity: they are converted from the CERs using the same Wang & Chen extinction-law family they then claim to confirm, so read the paper for the ratios, not the labels.\n\nWhat is actually new: first optical-to-12 micron extinction law spanning the whole Coalsack, with inner dense/diffuse splits and two reference regions. The pipeline is standard—blue-edge intrinsic colors, CE-CE linear fits—but executed cleanly: 368k Gaia-selected dwarfs in the intrinsic-colors region, binned median fitting with 3-sigma clipping, and external consistency checks against Guo et al. (2022) and Dobashi et al. (2005) that behave sensibly. The CER tables list sample sizes per band, which makes the sparse-region limits visible. That is a useful increment for dust studies, and the comparison with Wang et al. (2013) is fair and placed in context.\n\nThe soft spots, in order. The circularity I mentioned is the main one: Sections 3.3 and 4.4 adopt A_GBP/A_GRP from Wang & Chen (2019) and convert CERs to R_V with their R_V-dependent curves, so the stated agreement with R_V = 3.1 is partly built into the calibration. The quoted error bars come from the linear fits only; the C0 intrinsic-color calibration uncertainty is not propagated, and the authors themselves note the Teff-C0 fits diverge below 4500 K and above 7000 K. The MIR R_V = 5.5 conclusion is the least secure: W3 has only 229 sources, dense-region GLIMPSE samples are roughly 90–150, and high-extinction bins keep all points, so few outliers can set the slope. I checked the stress-test concern against the paper and it holds. Minor: the Coalsack sample size is 4,757 in the text but 14,112 in Table 1, presumably different cross-match stages, but it needs a sentence. The u-band deviation is handled honestly; it also hints the method's systematics are wavelength-dependent, which cuts toward caution in the MIR.\n\nWho this is for: anyone doing extinction or dust work in nearby clouds; the CER tables and map are citable. The R_V claims should be re-labeled as model-dependent in revision.\n\nRecommendation: send it to a serious referee. It deserves the referee time; the measurement is likely reliable and the target is well chosen, but the referee should insist on either propagating or explicitly bounding the C0 errors and softening the R_V=5.5 language.","headline":"Solid measurement paper whose headline R_V values are more model-dependent than the text admits; the color-excess ratios and extinction map are the real contribution.","tokens_in":20028,"tokens_out":3371,"would_cite":true,"duration_ms":32715,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The dust in the starless Coalsack cloud follows R_V = 3.1 in the optical and R_V = 5.5 in the mid-infrared.","keywords":["extinction law","interstellar dust","Coalsack molecular cloud","color excess ratio","reddening law","mid-infrared extinction","Gaia DR3 dwarf stars","R_V"],"falsifier":"Take a few hundred Coalsack-region dwarfs with independent high-resolution spectroscopy, assign each star its intrinsic color by spectral type, and recompute $E(B-V)$ and $R_V$ from the same photometry; if the resulting $R_V$ disagrees with $3.24 \\pm 0.32$ beyond the quoted uncertainties, the blue-edge calibration is the weak point. A second check: re-derive the extinction law using only stars whose line-of-sight reddening from an independent 3D dust map is consistent with zero and see whether the $R_V = 3.1$/5.5 split survives.","tokens_in":18842,"feed_emoji":"☁️","tokens_out":9579,"duration_ms":84107,"temperature":0.7,"pith_summary":"This paper tries to establish that the dust in the Coalsack, a nearby quiescent starless cloud, follows the Milky Way's standard diffuse extinction law in the optical and near-infrared ($R_V = 3.1$) but switches to the denser-cloud law in the mid-infrared ($R_V = 5.5$). It derives this from color excesses in 20 bands for 368,524 dwarf stars selected with Gaia DR3, using blue-edge $T_{\\rm eff}$--intrinsic-color relations to set the zero-reddening baseline and linear fits to color-excess diagrams for the ratios. The paper also finds that the optical law is nearly uniform across the cloud: for $E(B-V) > 0.3$ mag the average is $R_V = 3.24 \\pm 0.32$ with no significant $R_V$--$E(B-V)$ correlation, and it produces a $1.3'$ resolution reddening map that resolves fine cloud structure. If these results hold, a cloud with no star formation nevertheless shows the same environmental split in dust properties as active star-forming clouds, and the optical extinction curve can be treated as constant for reddening corrections in this region.","feed_headline":"Quiet Coalsack cloud follows R_V = 3.1 in optical, 5.5 in infrared","feed_subtitle":"A 368,524-star survey finds the starless cloud's dust matches both the diffuse Milky Way and star-forming clouds.","key_machinery":"The load-bearing mechanism is the blue-edge method for intrinsic colors: in a broad, mostly unreddened reference region (the Intrinsic Colors Region), the bluest 3% of stars in each 100-K $T_{\\rm eff}$ bin, restricted to $A_G < 0.05$ mag, are fitted with cubic polynomials to give $T_{\\rm eff}$--$C_0$ relations for all 20 bands; these relations convert observed colors into color excesses. The color-excess ratios $k_{\\lambda_1} = E_{GRP,\\lambda}/E_{GBP,GRP}$ and $k_{\\lambda_2} = E_{J,\\lambda}/E_{J,K_S}$ come from linear fits to CE--CE diagrams with 0.01-mag binning and iterative 3$\\sigma$ clipping, and are converted to $A_\\lambda/A_V$ and $A_\\lambda/A_{K_S}$ using the anchor ratios of the Wang & Chen (2019) extinction law.","core_discovery":"The central claim is that the multi-wavelength extinction law of the Coalsack is not a single curve: over $0.35$--$2.15\\,\\mu$m the color-excess ratios and relative extinction follow the $R_V = 3.1$ law of the diffuse Galactic interstellar medium, while over $2.15$--$12\\,\\mu$m the curve flattens and matches the $R_V = 5.5$ model of Weingartner & Draine (2001), the behavior previously seen in active star-forming clouds. Regional comparisons within the survey show the densest inner regions have the lowest $A_\\lambda/A_V$ and $A_\\lambda/A_{K_S}$ values, diffuse outer regions are flatter still, and the whole cloud sits in between; but despite these infrared variations, $R_V$ converted from color-excess ratios in $0.5^\\circ \\times 0.5^\\circ$ sub-regions shows no strong dependence on $E(B-V)$ above $0.3$ mag, with a Gaussian mean of $R_V = 3.24 \\pm 0.32$.","pith_inferences":["If the blue-edge calibration is unbiased, the same two-law pattern should appear in other quiescent, starless clouds; finding a quiescent cloud with an optical-NIR law different from $R_V = 3.1$ would show that environment, not star formation, controls the extinction law.","The weak spatial anti-correlation between $R_V$ and $E(B-V)$ below 0.3 mag could be a calibration artifact of the blue-edge method rather than a physical dust change; a spectroscopic sample in that low-extinction regime would separate the two.","Because the MIR law flattens further in the diffuse reference regions, the dust-grain population in the Coalsack's outer envelope appears at least as processed as in active star-forming clouds, a testable prediction for future observations of ice features and PAH emission.","Extending the same analysis to more distant or more embedded stars in the Coalsack would test whether the $R_V = 5.5$ flattening persists into even denser cores or gives way to steeper laws as grain growth proceeds."],"forward_implications":["For any object behind or within the Coalsack, optical-NIR reddening corrections can use the standard $R_V = 3.1$ law while MIR corrections should use the flatter $R_V = 5.5$ curve.","The absence of star formation in the Coalsack does not prevent the MIR extinction from being flat, so the $R_V = 5.5$ MIR law is not exclusive to active star-forming clouds.","The mean $R_V = 3.24 \\pm 0.32$ for $E(B-V) > 0.3$ mag means that within this cloud, treating $R_V$ as a constant introduces errors smaller than the quoted scatter.","The $1.3'$ $E(B-V)$ map, which agrees with an earlier catalog to 0.03 mag while revealing finer structure, can serve as a higher-resolution reddening reference for the Coalsack region."],"supporting_citations":[{"why":"Supplies the A_GBP/A_GRP anchor ratio and the R_V = 3.1 reference extinction curve used to convert color-excess ratios into relative extinction.","marker":"Wang & Chen (2019)"},{"why":"Provides the R_V = 5.5 model curve that the measured mid-infrared extinction is compared against and found to match.","marker":"Weingartner & Draine (2001)"},{"why":"Provides the R_V = 3.1 extinction-law curve used as the optical-NIR comparison standard.","marker":"Hensley & Draine (2020)"},{"why":"Earlier infrared extinction measurements in Coalsack sub-regions that this full-cloud study refines and compares against.","marker":"Wang et al. (2013)"},{"why":"The E(B-V) catalog whose values are compared to the new measurements to validate the extinction map.","marker":"Guo et al. (2022)"},{"why":"The A_V contour map used to define the Coalsack and reference regions and to provide comparative sub-region A_V values.","marker":"Dobashi et al. (2005)"},{"why":"Supplies the binning and iterative sigma-clipping method for linear fitting of CE-CE diagrams with inhomogeneous source densities.","marker":"Sun et al. (2021)"}],"fun_headline_variants":["Coalsack cloud's dust: R_V 3.1 in optical, 5.5 in infrared","Starless cloud Coalsack shows dual extinction law: 3.1 and 5.5","368K stars reveal Coalsack's split extinction law: 3.1 vs 5.5","Coalsack extinction: optical R_V 3.1, infrared R_V 5.5","Dust law in quiet Coalsack varies: optical 3.1, IR 5.5"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole result rests on the blue stars used as zero-reddening references being truly unreddened (the bluest 3% in a broad region, with $A_G < 0.05$ mag) and on the cubic fits to their colors giving the correct intrinsic color for every dwarf in the cloud; if those calibrators carry residual reddening, or the fit is biased for the reddened population, every color excess and every $R_V$ value shifts.","fun_headline_variants_meta":{"raw":{"variants":["Coalsack cloud's dust: R_V 3.1 in optical, 5.5 in infrared","Starless cloud Coalsack shows dual extinction law: 3.1 and 5.5","368K stars reveal Coalsack's split extinction law: 3.1 vs 5.5","Coalsack extinction: optical R_V 3.1, infrared R_V 5.5","Dust law in quiet Coalsack varies: optical 3.1, IR 5.5"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000899,"raw_usage":{"total_tokens":3946,"prompt_tokens":1093,"completion_tokens":2853,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":709,"completion_tokens_details":{"reasoning_tokens":2723}},"tokens_in":709,"tokens_out":2853,"duration_ms":18799,"temperature":1.0,"reasoning_tokens":2723,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T23:06:41.658980+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a few hundred Coalsack-region dwarfs with independent high-resolution spectroscopy, assign each star its intrinsic color by spectral type, and recompute $E(B-V)$ and $R_V$ from the same photometry; if the resulting $R_V$ disagrees with $3.24 \\pm 0.32$ beyond the quoted uncertainties, the blue-edge calibration is the weak point. A second check: re-derive the extinction law using only stars whose line-of-sight reddening from an independent 3D dust map is consistent with zero and see whether the $R_V = 3.1$/5.5 split survives.","supporting_citations":[],"review_version":1}