Pith. sign in

REVIEW 3 major objections 5 minor 46 references

A Detailed Analysis of Intermediate-scale Structure in Optical Extinction Curves: Expanded Census and Two-family Classification

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A systematic census of 74 Milky Way sightlines reports nine new intermediate-scale dust extinction features and shows that the full set sorts into two families, α and β, with the α family strongly correlated with the 2175 Å carbon bump.

desk verdict A well-documented, honest census of intermediate-scale extinction features; the nine new candidates are plausible but need independent confirmation. read the letter →

arxiv 2608.06611 v1 pith:R3WXKVO3 submitted 2026-08-06 astro-ph.GA

C. M. Gunasekera , M. Decleir , K. D. Gordon , G. C. Clayton , B. Günay , D. V. De Putte , P. Yanchulova This is my paper · ORCID
classification astro-ph.GA
keywords interstellardustextinctionintermediate-scalestructure2175ÅbumpdiffusebandsDrudeprofilecarbonaceousHubbleSpaceTelescopeSTISMilkyWaysightlines
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper reports a systematic census of intermediate-scale structure (ISS) in the optical extinction curves of 74 Milky Way sightlines, combining 24 new HST/STIS spectra with 50 literature targets. It finds that the Milky Way's extinction curve contains far more broad structure than previously known: nine new candidate features beyond the six already reported, with fifteen of the seventeen total features detected above 5σ. The paper's central interpretive claim is that these features sort into two families, α and β, according to how their strengths correlate across sightlines. Family α sits at shorter wavelengths and tracks the strength of the 2175 Å bump and the dust size parameter 1/R(55); family β sits at longer wavelengths and correlates more weakly with the bump. If correct, this means the carriers of these broad extinction features form two distinct, possibly competing carbonaceous dust populations, and the strength of the 2175 Å bump can be used to predict part of the optical extinction structure.

What carries the argument

The analysis rests on the Drude profile as the model for each ISS feature, $D(x,x_0,\gamma)=x^2\gamma^2/((x^2-x_0^2)^2+(x\gamma)^2)$ in inverse-micron wavenumber $x=1/\lambda$, added to a fourth-order polynomial that represents the optical continuum. The measurement pipeline fits each sightline's extinction curve with this sum, then examines the residuals after removing known interstellar lines and diffuse interstellar bands and convolving with a Gaussian that suppresses structure narrower than 150 Å. The mean residual across 74 sightlines is used to iteratively add Drude components until the residual is flat; the statistical significance of each feature is computed from the area under its Drude profile relative to the standard deviation of the mean. This combination of a fixed functional form for the broad features and an ensemble average over many independent lines of sight is what lets the paper claim detections of features whose individual amplitudes are only a few percent of A(55).

What would settle it

Re-fit the same 74 extinction curves with a differently shaped continuum (for instance a spline or a fifth- or sixth-order polynomial) and with an independent set of stellar atmosphere models, then check whether the nine new candidate features still appear at >5σ in the mean residual; if they disappear or drop below 5σ under a reasonable alternative continuum, the claim that they are real interstellar features would be falsified. A complementary test is to look for the same features in higher signal-to-noise spectra of individual sightlines, or to require that the candidate features correlate with mid-infrared carbonaceous emission or absorption features as the carbon-origin claim predicts.

Watch

Extended reading notes

Core claim

The authors establish, on the basis of a homogeneous fit of 74 extinction curves, that the optical spectrum of interstellar extinction contains at least 17 broad (FWHM ≳ 150 Å) features with widths larger than any known diffuse interstellar band. Six of these had been reported before; this work confirms five of them, leaves the 5400 Å feature unconfirmed at 3.2σ, and proposes nine new candidates that survive both a 5σ detection threshold and a check that their strengths rise with dust column A(55). The central new result is a two-family classification: correlation analysis of the feature amplitudes splits the sample into family α (ISS32, ISS36, ISS40, ISS43, ISS44, ISS48, ISS52, ISS54) and family β (ISS59, ISS62, ISS64, ISS69, ISS75, ISS77, ISS81, ISS84), with ISS96 in neither family. Members of family α strongly track each other and the 2175 Å bump strength; family β members track each other and show moderate bump correlations, while ISS36 is anti-correlated with several β features, indicating two carrier populations that can be enhanced or suppressed relative to one another. The paper argues that the correlations with the 2175 Å bump point to a carbonaceous origin for the carriers, with family α's additional correlation with 1/R(55) suggesting a size-sensitive carbonaceous population and family β a size-insensitive one.

Load-bearing premise

The whole detection scheme assumes that any residual larger than 5σ left after subtracting a fourth-order polynomial continuum plus known Drude features is a real astrophysical feature, rather than an artifact of the continuum model, the stellar atmosphere models, or the fringe correction; the paper itself warns that if the literature features are not exactly Drude-shaped, some of the new candidates could be artifacts.

Editorial extensions

If this is right

  • The optical extinction curve of the Milky Way is not smooth: it contains at least fifteen statistically significant broad features between roughly 3200 and 9600 Å, nine of them newly reported here.
  • The two families give a new organizing axis for extinction-curve studies: a sightline's family-α to family-β strength ratio encodes which dust population dominates, allowing extinction curves to be classified by carrier family rather than only by R(55).
  • The strong correlation of ISS43, ISS48, ISS64 and ISS77 with the 2175 Å bump means the bump strength can be used to predict part of the optical extinction structure, and it ties these four features to carbonaceous carriers.
  • ISS54 (5400 Å) is not confirmed by this sample, and ISS62 is only marginally detected, so those two features need higher signal-to-noise data before they can be treated as established.
  • The anti-correlation of ISS36 with family-β features implies the two carrier populations are not simply independent but can be competitively enhanced or suppressed in different environments.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the two-family pattern survives in the Magellanic Clouds, the wavelength split between α and β may map onto a difference in electronic transition energies of two carbonaceous grain populations, which would make the α/β ratio a metallicity-sensitive diagnostic.
  • The authors' comparison of ISS with DIBs suggests a testable extension: a dedicated cross-correlation of ISS family strengths with DIB equivalent widths along the same 74 sightlines would show whether the two are different manifestations of the same carbonaceous carrier population at different size scales.
  • Because family α tracks 1/R(55) while family β does not, one could use the α/β amplitude ratio as a new observational handle on grain size distributions, independent of the standard R(V) parametrization.
  • A direct laboratory test: if the carriers are carbonaceous, the features' central wavelengths should match electronic transitions of specific PAH cations or hydrogenated amorphous carbon clusters, which could be checked against published laboratory spectra.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. This paper presents a homogeneous analysis of intermediate-scale structure (ISS) in optical extinction curves using 74 lines of sight, combining 50 literature sightlines with 24 new HST/STIS observations. The authors fit each extinction curve with a fourth-order polynomial plus Drude profiles for known ISS features, then iteratively add Drude profiles to the mean residual until it is flat. They confirm five of six literature ISS features at >5σ, fail to confirm ISS54, and report ten new candidate features (later reduced to nine after excluding ISS96). Based on Pearson correlations between fitted amplitudes, they group most features into two families, α and β, and report strong correlations of ISS43, ISS48, ISS64, and ISS77 with the 2175 Å bump strength, which they interpret as evidence for carbonaceous carriers.

Significance. If the candidate features are real, this work substantially expands the census of broad extinction structure in the Milky Way and provides a new phenomenological organization (two families) that could constrain dust models. The paper's strengths include the new STIS data, the homogeneous re-analysis of the combined sample, the public release of code and data, and an explicit statement of the main assumption (Drude representation of literature features) in Section 7. However, the detection significance and the correlation-based family classification rest on several methodological choices that need additional validation before the central claims can be accepted.

major comments (3)
  1. [4.3, Table 5, Section 7] The significance estimates for the candidate ISS features are computed by MCMC-fitting the same mean residual from which the features were selected, using the standard deviation of the mean as the uncertainty. This is a post-hoc selection: a feature that appears in the residual is then tested against that same residual, so the quoted 5σ values do not include the multiplicity of trials or the uncertainty in the continuum model. The paper's own Section 7 states that the decomposition assumes the six literature features are well-modelled by Drude profiles and that 'some of the proposed candidate ISS features may be artifacts of the method' if this is not the case. Because the discovery claim of nine new features depends directly on this procedure, the authors should provide robustness tests, e.g., fitting with a 5th-order polynomial or an alternative continuum model, a bootstrap or false-discovery-rate analysis, and split-sample validation, to demonstrate that the candidates are not absorbing continuum mis-specification.
  2. [5, Figure 10, Figure 11] The two-family classification is derived entirely from Pearson correlation coefficients between fitted ISS amplitudes, but the paper does not give uncertainties on r, does not state significance thresholds, and does not explain the clustering rule used to assign features to families. Several low-significance features (ISS32, ISS54, ISS62) are tentatively placed in families, while ISS96 is excluded on the basis of a lack of correlation. The authors should propagate the amplitude uncertainties into the correlation analysis (e.g., via bootstrap or Monte Carlo), test the significance of the family separation with a permutation test or a clustering algorithm, and state the criterion used to assign a feature to a family. Without this, the two-family classification is not quantitatively supported.
  3. [6, Figure 11] The claim of strong correlations between the 2175 Å bump strength B3 and ISS43, ISS48, ISS64, and ISS77 is based on Pearson r values without uncertainties. Because B3 and the ISS amplitudes are derived from fits to the same extinction curves and are normalized by A(55), possible correlated errors or the shared normalization could bias r upward. The paper should report p-values or bootstrap confidence intervals for these correlations and ideally use a regression method that accounts for errors in both variables before concluding a carbonaceous origin.
minor comments (5)
  1. [4.2, Table 5, Figure 11] The text says ISS54 cannot be confirmed, yet Table 5 lists it as a feature and Figure 11 places it in Family α; please clarify whether it is considered a detection or a tentative feature.
  2. [3.3, Eq. (4)] The statement that a fourth-order polynomial is adequate is taken from M20; please state whether this was re-verified for the present 74-line-of-sight sample, given the extended wavelength range to 10000 Å.
  3. [Figure 7] The panel labels such as 'model = 4D Poly. + 3 Drudes' are inside the figure and not defined in the caption; define the abbreviations in the caption.
  4. [Table 5] Entries like 'Yesnew', 'Nonew', 'Unsurenew' in the 'Dust feature?' column are cryptic; use separate columns for 'Literature?' and 'Dust feature?' with a clear key.
  5. [2.1] The informal naming of the sample and the footnote about Prince, while harmless, is out of keeping with the tone of an ApJ paper; consider a more standard designation.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the ISS census, family classification, and bump correlations are empirical fits rather than derivations that reduce to their inputs; the only caution is that candidate significances are computed from the same residuals used to select the features.

full rationale

The paper's derivation chain is: (1) fit TLUSTY model atmospheres and measure extinction curves; (2) model optical extinction as a fourth-order polynomial plus Drude profiles (Eq. 4); (3) inspect mean residuals, iteratively add Drude profiles until the residual is flat (Section 4.3); (4) compute significances by MCMC-fitting all 17 features to the mean residual (Table 5); and (5) correlate the fitted amplitudes. None of these steps defines an ISS feature in terms of its own detection, nor fits a parameter to a subset and then presents the same quantity as an independent prediction. The candidate features are selected from the residuals, and their significances are then estimated from those same residuals; this is a post-selection or overfitting concern that can inflate significances if the fourth-order continuum or Drude model is imperfect, and the paper explicitly concedes in Section 7 that if the literature features are not well modelled by Drudes, 'some of the proposed candidate ISS features may be artifacts of the method.' That is a correct statement of model dependence, not a tautology. The B3-ISS correlations are between separately fitted parameters (the UV bump strength from Eq. 6 and optical Drude amplitudes from Eq. 4) that share only the A(55) normalization; any induced correlation is a covariance artifact, not an identity by construction. Citations to M20 and to Gordon et al. provide prior methodology and empirical calibration rather than a uniqueness theorem invoked to forbid alternatives. In short, the paper is self-contained against external data; no claimed result reduces to its inputs by definition.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper adds no new physical entities. Its central claims rest on the chosen continuum model (fourth-order polynomial), the Drude shape for features, the 150 Å width filter, and the stellar atmosphere models used to derive extinction curves. These are domain assumptions from the literature. The fitted feature parameters and the hand-chosen width cutoff are the main free parameters.

free parameters (4)
  • ISS feature Drude parameters (center, width, amplitude) for 17 features = Table 5 (e.g., ISS43: 4353.5 Å, 504.1 Å, amplitude 364.1e-4)
    These fitted values define the new candidates and drive the family correlations; they are fit to each sightline's extinction curve.
  • Fourth-order polynomial continuum coefficients E_j^{A(55)} = Table 3 per sightline
    The polynomial baseline determines what appears as residual features; changing the order changes the derived ISS amplitudes.
  • Gaussian smoothing width = 150 Å
    Hand-chosen cutoff to suppress DIBs; sets the minimum width of structures classified as ISS.
  • A(55) extinction normalization = Table 3 per sightline
    Normalization derived from JHK photometry; errors in A(55) propagate to all normalized amplitudes, including B3, potentially inflating correlations.
assumptions (5)
  • domain assumption The optical extinction curve is well described by a fourth-order polynomial plus Drude profiles.
    Used in Eq. (4); if the true continuum requires a higher-order polynomial, residual features could be artifacts.
  • domain assumption TLUSTY non-LTE stellar atmosphere models accurately represent the intrinsic SEDs of the OB targets.
    Section 3.1 fits each star with these models; model errors propagate into the extinction curves and residuals.
  • domain assumption All ISS features have Drude (damped harmonic oscillator) shapes.
    Eq. (5); if real feature shapes differ, the fitted amplitudes and widths could be biased.
  • domain assumption A Gaussian convolution with 150 Å width separates ISS from DIBs.
    Section 4 uses this filter; features narrower than 150 Å are excluded by construction.
  • domain assumption The UV and far-UV extinction curves contain no ISS features.
    Section 4 justifies limiting the analysis to optical based on prior work and this paper's own data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Detailed Analysis of Intermediate-scale Structure in Optical Extinction Curves: Expanded Census and Two-family Classification." pith.science (2026). https://pith.science/paper/R3WXKVO3

@misc{pith2026260806611,
  author       = {Pith},
  title        = {Pith review of: A Detailed Analysis of Intermediate-scale Structure in Optical Extinction Curves: Expanded Census and Two-family Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R3WXKVO3}},
  note         = {Machine review of arXiv:2608.06611}
}
abstract

The features of interstellar extinction curves serve as powerful diagnostics for interstellar dust, revealing information about its composition, size distribution, and the physical and chemical processes that shape it. D. Massa et al. reported three faint but wide extinction features, termed intermediate-scale structures (ISS) at 4370, 4870, 6300 \r{A}. Since then, three additional ISS features have been reported in the literature at 7700, 5400, 8500 \r{A}, all with widths greater than any known diffuse interstellar band. We present new optical and UV Hubble Space Telescope/STIS spectra for a sample of 24 early-type OB stars. We used these data combined with 50 literature targets for a systematic, homogeneous analysis of ISS features, with the aim of investigating their observational behavior to help constrain their carriers. This analysis revealed nine more candidate ISS features. We also find that ISS features can be arranged into two main families, which we call $\alpha$ and $\beta$, according to correlations between the feature strengths. Finally, we find strong correlations between the 2175 \r{A} bump strength and ISS features at 4353, 4847, 6443, 7710 \r{A}, suggesting their possible carbonaceous origin.

Figures

Figures reproduced from arXiv: 2608.06611 by the authors.

Figure 1
Figure 1. Optical and UV spectra of the 24 stars in the Prince sample. The HST/STIS spectra are shown as solid lines, and the photometric data (U, B, V, J, H, K) are shown as open circles. The blue and green colors are simply used to distinguish between adjacent curves, and do not carry any particular meaning. The spectra are roughly ordered according to the size of the UV bump. et al. 1992), by comparing the spectral energy … view at source ↗
Figure 2
Figure 2. HST/STIS spectrum (black) of ALS6672, a B1V type star, fit with measure extinction package and TLUSTY using MCMC. The unreddened best-fit stellar atmosphere model is in blue, and the reddened model is in red. The three plot points on the right are the J, H, and K photometric bands. best-fit stellar model. Then we convert E(λ − 55) to the absolute extinction A(λ) normalized by the value at 5500 ˚A using the relation,… view at source ↗
Figure 3
Figure 3. A comparison plot between the log Teff , and log(g) derived in F19 using spline interpolation, and those derived in this paper. ing the first two significant figures of their central wave￾length. For reference, the mean residual from the 4D polynomial-only fit is shown in the background using a dashed line with reduced opacity. We shade in five times the standard deviation of the mean of the resid￾uals, showing that… view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Measured UV and optical extinction curves of the Prince sample. The curves are ordered by the overall slope of the extinction curve. The solid lines show the extinction curve spectral data, while the open circles give the JHK photometric points of the curve. The blue a…
Figure 5
Figure 5. Figure 5: Measured (black) and fitted (red and blue) extinction curve for ALS6672, using the FM90 parameterization in the UV, and a 4D polynomial with Drude profiles in the optical. The “x” points represent outlying points excluded from the fit. The residuals are shown in the bo…
Figure 6
Figure 6. Figure 6: Residuals of fitting the optical extinction curves. Top: The blue dashed plots show the residuals of all 74 targets at the full resolution. The black solid plot is the mean of the residuals convolved with a Gaussian filter, resolving only features wider than 150 ˚A. Th…
Figure 7
Figure 7. Figure 7: Average of the residuals of the extinction curves, fitted with the 2 literature-found (top panel) and 11 new (bottom two panels) candidate ISS features. The solid lines represent the averaged residuals resulting from fitting the indicated ISS features. For comparison, …
Figure 8
Figure 8. Figure 8: The average residual of fitting the ISS feature at 5400 ˚A feature in orange. For comparison we include the same residual appearing at the bottom panel of [PITH_FULL_IMAGE:figures/full_fig_p015_8.png]
Figure 9
Figure 9. Figure 9: Example plots of amplitude of ISS features, plotted against total extinction, A(55), (top row) and 2175 ˚A UV bump amplitude, B3 (middle row), and other ISS feature amplitudes (bottom row). Note that the y-axis in the top row is the un-normalized feature strength A(ISS…
Figure 10
Figure 10. Figure 10: Colormap of the correlation coefficients between the feature strengths with A(55), 1/R(55), the bump strength (B3), the far-UV rise (C A(55) 4 ), log of the effective stellar temperature (log Teff ), hydrogen column density f(H2) and each other. The α and β superscrip…
Figure 11
Figure 11. Figure 11: Circle graph of correlations between the 17 published and candidate ISS features. We present here only the moderate and strong correlations, grouping the two main ISS families. The lines represented the correlation coefficient between two ISS amplitudes, and B A(55) 3…
Figure 12
Figure 12. Figure 12: Top: The ISS features based on their fitted parameters. The families are indicated by color and linestyle. The features with label “ISSnn*” indicate features that have correlations with other members in the family but may not be dust features or have a low detection. …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 15 canonical work pages

  1. [1]

    P., Tollerud, E

    Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33, doi: 10.1051/0004-6361/201322068 Astropy Collaboration, Price-Whelan, A. M., Sip˝ ocz, B. M., et al. 2018, AJ, 156, 123, doi: 10.3847/1538-3881/aabc4f Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., et al. 2022, ApJ, 935, 167, doi: 10.3847/1538-4357/ac7c74

    doi:10.1051/0004-6361/201322068 2013

  2. [2]

    Prochaska, J. X. 2015, MNRAS, 447, 545, doi: 10.1093/mnras/stu2448

    work page doi:10.1093/mnras/stu2448 2015

  3. [3]

    R., & Robinson, D

    Bevington, P. R., & Robinson, D. K. 2003, Data reduction and error analysis for the physical sciences

    2003

  4. [4]

    C., Gordon, K

    Bohlin, R. C., Gordon, K. D., & Tremblay, P.-E. 2014, PASP, 126, 711, doi: 10.1086/677655

    doi:10.1086/677655 2014

  5. [5]

    A., Clayton, G

    Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245, doi: 10.1086/167900

    doi:10.1086/167900 1989

  6. [6]

    A., Sembach, K

    Cardelli, J. A., Sembach, K. R., & Mathis, J. S. 1992, AJ, 104, 1916, doi: 10.1086/116367

    work page doi:10.1086/116367 1992

  7. [7]

    C., Gordon, K

    Clayton, G. C., Gordon, K. D., Salama, F., et al. 2003, ApJ, 592, 947, doi: 10.1086/375771

    work page doi:10.1086/375771 2003

  8. [8]

    D., Andrews, J

    Decleir, M., Gordon, K. D., Andrews, J. E., et al. 2022, ApJ, 930, 15, doi: 10.3847/1538-4357/ac5dbe

    doi:10.3847/1538-4357/ac5dbe 2022

Show all 46 references
  1. [9]

    Draine, B. T. 2003, ARA&A, 41, 241, doi: 10.1146/annurev.astro.41.011802.094840

    2003 arXiv

  2. [10]

    Draine, B. T. 2011, Physics of the Interstellar and Intergalactic Medium (Princeton University Press)

    2011

  3. [11]

    L., & Massa, D

    Fitzpatrick, E. L., & Massa, D. 1990, ApJS, 72, 163, doi: 10.1086/191413

    1990 doi

  4. [12]

    L., & Massa, D

    Fitzpatrick, E. L., & Massa, D. 2005, AJ, 130, 1127, doi: 10.1086/431900

    2005 doi

  5. [13]

    Clayton, G. C. 2019, ApJ, 886, 108, doi: 10.3847/1538-4357/ab4c3a

    2019 doi

  6. [14]

    W., Lang, D., & Goodman, J

    Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP, 125, 306, doi: 10.1086/670067 22

    2013 doi

  7. [15]

    2020, AJ, 159, 113, doi: 10.3847/1538-3881/ab6d01

    Galazutdinov, G., Bondar, A., Lee, B.-C., et al. 2020, AJ, 159, 113, doi: 10.3847/1538-3881/ab6d01

    2020 doi

  8. [16]

    2024, The Journal of Open Source Software, 9, 7023, doi: 10.21105/joss.07023

    Gordon, K. 2024, The Journal of Open Source Software, 9, 7023, doi: 10.21105/joss.07023

    2024 doi

  9. [17]

    Gordon, K., Decleir, M., & Gunasekera, C. M. 2025, measure extinction: M31 UV Extinction paper release, v1.4 Zenodo, doi: 10.5281/zenodo.15831932

    2025 doi

  10. [18]

    D., Cartledge, S., & Clayton, G

    Gordon, K. D., Cartledge, S., & Clayton, G. C. 2009, ApJ, 705, 1320, doi: 10.1088/0004-637X/705/2/1320

    2009 doi

  11. [19]

    D., Clayton, G

    Gordon, K. D., Clayton, G. C., Decleir, M., et al. 2023, ApJ, 950, 86, doi: 10.3847/1538-4357/accb59

    2023 doi

  12. [20]

    D., Misselt, K

    Gordon, K. D., Misselt, K. A., Bouwman, J., et al. 2021, ApJ, 916, 33, doi: 10.3847/1538-4357/ac00b7

    2021 doi

  13. [21]

    D., Fitzpatrick, E

    Gordon, K. D., Fitzpatrick, E. L., Massa, D., et al. 2024, ApJ, 970, 51, doi: 10.3847/1538-4357/ad4be1

    2024 doi

  14. [22]

    M., Zhang, X., & Zhang, R

    Green, G. M., Zhang, X., & Zhang, R. 2025, ApJ, 988, 5, doi: 10.3847/1538-4357/addd11

    2025 doi

  15. [23]

    2026, HST/STIS UV and Optical Spectrophotometric data for the Prince Sample of Milky Way sightlines, Zenodo, doi: 10.5281/zenodo.19665233

    Gunasekera, C., Decleir, M., Gordon, K., et al. 2026, HST/STIS UV and Optical Spectrophotometric data for the Prince Sample of Milky Way sightlines, Zenodo, doi: 10.5281/zenodo.19665233

    2026 doi

  16. [24]

    Gunasekera, C. M. 2026, Chamani8/hst iss ext: hst iss ext v1, v1 Zenodo, doi: 10.5281/zenodo.20274225

    2026 doi

  17. [25]

    R., Millman, K

    Harris, C. R., Millman, K. J., van der Walt, S. J., et al. 2020, Nature, 585, 357, doi: 10.1038/s41586-020-2649-2

    2020 doi

  18. [26]

    Herbig, G. H. 1963, ApJ, 137, 200, doi: 10.1086/147496

    1963 doi

  19. [27]

    Herbig, G. H. 1995, ARA&A, 33, 19, doi: 10.1146/annurev.aa.33.090195.000315

    1995

  20. [28]

    2025, AJ, 169, 178, doi: 10.3847/1538-3881/adb1bf

    Hubeny, I., Bohlin, R., Gordon, K., & Fitzpatrick, E. 2025, AJ, 169, 178, doi: 10.3847/1538-3881/adb1bf

    2025 doi

  21. [29]

    Hunter, J. D. 2007, Computing in Science & Engineering, 9, 90, doi: 10.1109/MCSE.2007.55

    2007 doi

  22. [30]

    P., D’Hendecourt, L., & L´ eger, A

    Joblin, C., Maillard, J. P., D’Hendecourt, L., & L´ eger, A. 1990, Nature, 346, 729, doi: 10.1038/346729a0

    1990 doi

  23. [31]

    A., Whittet, D

    Larson, K. A., Whittet, D. C. B., & Hough, J. H. 1996, ApJ, 472, 755, doi: 10.1086/178104

    1996 doi

  24. [32]

    Li, A., & Draine, B. T. 2001, ApJ, 554, 778, doi: 10.1086/323147 Ma ´ ız Apell´ aniz, J., Barb´ a, R. H., Caballero, J. A., Bohlin, R. C., & Fari˜ na, C. 2021, MNRAS, 501, 2487, doi: 10.1093/mnras/staa2371

    2001 doi

  25. [33]

    L., & Gordon, K

    Massa, D., Fitzpatrick, E. L., & Gordon, K. D. 2020, ApJ, 891, 67, doi: 10.3847/1538-4357/ab6f01

    2020 doi

  26. [34]

    D., & Fitzpatrick, E

    Massa, D., Savage, B. D., & Fitzpatrick, E. L. 1983, ApJ, 266, 662, doi: 10.1086/160813 Pantaleoni Gonz´ alez, M., Ma ´ ız Apell´ aniz, J., Barb´ a, R. H., & Reed, B. C. 2021, MNRAS, 504, 2968, doi: 10.1093/mnras/stab688

    1983 doi

  27. [35]

    2024, STIS Data Handbook, Version 8.0, Space Telescope Science Institute, Baltimore, MD

    Rickman, E., Brown, J., et al. 2024, STIS Data Handbook, Version 8.0, Space Telescope Science Institute, Baltimore, MD. https://hst-docs.stsci.edu/stisdhb

    2024

  28. [36]

    K., & Green, G

    Saydjari, A. K., & Green, G. M. 2025, arXiv e-prints, arXiv:2507.07162, doi: 10.48550/arXiv.2507.07162

    2025 doi arXiv

  29. [37]

    P., & McCall, B

    Snow, T. P., & McCall, B. J. 2006, ARA&A, 44, 367, doi: 10.1146/annurev.astro.43.072103.150624

    2006 arXiv

  30. [38]

    P., & Donn, B

    Stecher, T. P., & Donn, B. 1965, ApJ, 142, 1681, doi: 10.1086/148461

    1965 doi

  31. [39]

    2011, ApJ, 742, 2, doi: 10.1088/0004-637X/742/1/2

    Steglich, M., Bouwman, J., Huisken, F., & Henning, T. 2011, ApJ, 742, 2, doi: 10.1088/0004-637X/742/1/2

    2011 doi

  32. [40]

    Tielens, A. G. G. M. 2005, The Physics and Chemistry of the Interstellar Medium

    2005

  33. [41]

    A., Clayton, G

    Valencic, L. A., Clayton, G. C., & Gordon, K. D. 2004, ApJ, 616, 912, doi: 10.1086/424922 Van De Putte, D., Cartledge, S. I. B., Gordon, K. D.,

    2004 doi

  34. [42]

    C., & Roman-Duval, J

    Clayton, G. C., & Roman-Duval, J. 2023, ApJ, 944, 33, doi: 10.3847/1538-4357/ac9902 Van Rossum, G., & Drake, F. L. 2009, Python 3 Reference Manual (Scotts Valley, CA: CreateSpace)

    2023 doi

  35. [43]

    E., et al

    Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261, doi: 10.1038/s41592-019-0686-2

    2020 doi

  36. [44]

    C., & Draine, B

    Weingartner, J. C., & Draine, B. T. 2001, ApJ, 548, 296, doi: 10.1086/318651

    2001 doi

  37. [45]

    Whiteoak, J. B. 1966, ApJ, 144, 305, doi: 10.1086/148605

    1966 doi

  38. [46]

    2024, ApJ, 971, 127, doi: 10.3847/1538-4357/ad613e

    Zhang, R., Yuan, H., Huang, B., et al. 2024, ApJ, 971, 127, doi: 10.3847/1538-4357/ad613e

    2024 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.