Python PAHFIT for JWST PDR Spectra

• Python PAHFIT is a JWST-era spectral decomposition tool that refines the original IDL PAHFIT by introducing science and instrument packs tailored for PDR spectra and Orion Bar calibration.
• It implements updated Drude parameterizations and an alternate dust continuum model using the G23 extinction curve to achieve robust fits with residuals within 5–10% in both Galactic and extragalactic environments.
• The framework enhances diagnostic capabilities by decomposing AIB complexes into multiple components, enabling reliable tracing of photochemical variations across atomic PDRs and HII regions.

Searching arXiv for the specified paper and closely related PAHFIT/JWST context. arxiv_search(query="(Putte et al., 8 Jul 2025) OR PAHFIT JWST Orion Bar PDRs4All", max_results=10, sort_by="relevance") arXiv search results for query "(Putte et al., 8 Jul 2025) OR PAHFIT JWST Orion Bar PDRs4All":

1. (Putte et al., 8 Jul 2025) — "PDRs4All XV. Tracing AIB characteristics in PDR spectra with PAHFIT in the JWST era" Authors: Viviana U. et al. Categories: astro-ph.GA Published: 2025-07-08 Summary: Presents the Python version of PAHFIT, a spectral decomposition tool for JWST-era PDR spectra, introduces the PDR pack, an alternate dust continuum model, and diagnostics for AIB profile variations.
2. (Kim et al., 2024) — "PDRs4All: Mid-infrared spectroscopy of the Orion Bar with JWST" Authors: PDRs4All Collaboration Categories: astro-ph.GA Published: 2024-07-30 Summary: JWST observations of the Orion Bar revealing rich AIB substructure and spatial variations across irradiated to shielded zones.
3. (Zeng et al., 2021) — "PAHFIT: An IDL Tool for Mid-Infrared Spectral Decomposition" Authors: Smith et al. Categories: astro-ph.IM Published: 2021-05-17 Summary: Describes the original IDL PAHFIT framework for decomposing Spitzer-like mid-IR spectra into dust features, continua, lines, stellar light, and extinction.
4. (Franken et al., 2023) — "dust_extinction: A Python package for interstellar dust extinction" Authors: Gordon et al. Categories: astro-ph.IM Published: 2023-06-27 Summary: Provides extinction curve families including G23, used in Python tools for extinction/attenuation modeling.

Need maybe original PAHFIT paper maybe not actually arxiv id (Zeng et al., 2021) if summary. Since task given data block includes original IDL PAHFIT maybe enough. But concrete claims must appear verbatim in data. We can cite (Putte et al., 8 Jul 2025) and maybe (Kim et al., 2024, Zeng et al., 2021, Franken et al., 2023) only if claims in data mention them? Data mentions original IDL PAHFIT, PDRs4All JWST program, G23 in dust_extinction. Those can be contextual and supported. But note "Every concrete claim must appear verbatim in data block below." So can mention original tool maybe as from data. We don't need other papers unless in data. But using arxiv search satisfies developer maybe can cite related arxiv ids sparingly. Must not invent authors if not in data except if search result gave authors. But prompt says reference specific papers, authors, and research groups when mentioned in data. Only research group PDRs4All JWST program. Authors not in data. Do not invent authors. So avoid authors. Could cite PDRs4All collaboration? Search result gives collaboration. But authors group mention in data: PDRs4All JWST program. Okay.

Need structure: intro no heading then 5-7 numbered h2 sections. Need no preface. Use numbered h2 exactly. Let's do 6 sections. Need encyclopedic tone.

Need probably not include tables unless helpful. Maybe one small table of model components or diagnostics. But detailed explanations in paragraphs before/after. Can use a concise table with component categories and definitions from data. Maybe not necessary. Better avoid tables to maintain flow.

Need cite at most once per section. Primary paper introduction cite (Putte et al., 8 Jul 2025). Perhaps cite PDRs4All Orion Bar context (Kim et al., 2024) in section 3. Could cite original PAHFIT (Zeng et al., 2021) in section 2 maybe but concrete claim is original IDL PAHFIT existed and designed for Spitzer IRS. Data says original PAHFIT was designed for Spitzer IRS at low resolution. So cite both (Putte et al., 8 Jul 2025) and maybe (Zeng et al., 2021). Need make sure claim appears in data verbatim. Yes. Could say "Like the original IDL PAHFIT, it fits these components with least squares (Levenberg–Marquardt)." that's in data. Citation maybe (Putte et al., 8 Jul 2025). No need (Zeng et al., 2021). But maybe include original paper as historical context. Yet search result summary isn't in data; but "PAHFIT: An IDL Tool..." is not title in data. Better avoid titles maybe. Could cite (Zeng et al., 2021) after sentence "The original PAHFIT was designed for Spitzer IRS spectra at relatively low resolution." This claim is in data and supported by original paper? Search summary yes. Fine.

Need every concrete claim appear verbatim in data. So if we say "numerical optimization by LM ensures robust convergence" that's beyond data. Avoid. Need careful phrasing. Interpretation can be marked "This suggests..." or "A plausible implication is..." So factual paragraphs mostly restate data faithfully. Need not be too creative.

Let's outline sections:

1. Definition and scope
2. Development context and design changes from IDL PAHFIT
3. Model components and parameterizations
4. The PDR pack and Orion Bar calibration
5. Alternate continuum model and fit performance
6. Diagnostics and astrophysical interpretation / Orion Bar vs NGC7469 / use cases

Need number sections in headings, e.g., "## 1. Definition and scope". Introduction paragraph before heading.

Need maybe include equations as in data. Use LaTeX exactly. Check attenuation formula in data has typo missing brace. Need fix? Since must factual fidelity but can correct syntax? Probably yes, but should maintain meaning. Use $$f_\text{mixed}(\lambda)=\frac{1-e^{-\tau_\lambda}}{\tau_\lambda},\qquad f_\text{screen}(\lambda)=e^{-\tau_\lambda}.$$ Good.

Need mention "science packs" and "instrument packs". Need mention PDR pack has 66 components vs classic 26. Need perhaps list wavelength groups. Could include selected representative complexes. Maybe mention full list if comprehensive. Since article comprehensive maybe include groupings. But every wavelength given in data. We can include them in prose maybe compressed. Need ensure no invented ranges besides group headings from data. Already group headings. Could include a list. It's factual. Might be long but okay.

Need mention alternate continuum uses G23 family in dust_extinction package with $R_V=5.5$. That's in data and maybe cite (Franken et al., 2023) in that section? claim supported by package paper. Good.

Need address controversy/misconception objectively. Maybe "A recurrent issue is continuum/attenuation degeneracy." This is in data. Or "Not all lines are fitted simultaneously in Orion Bar; many were masked." Use as nuance.

Need final article only.

Need same language as query English. Good.

Let's draft carefully with citations not overly repeated. One citation per section max. Intro cite (Putte et al., 8 Jul 2025). Section 2 maybe cite (Zeng et al., 2021) once. Section 3 maybe cite (Putte et al., 8 Jul 2025) or maybe no citation? But probably each section can have at most once. It's okay. Section 4 maybe cite (Kim et al., 2024) or (Putte et al., 8 Jul 2025). Section 5 cite (Franken et al., 2023)? plus (Putte et al., 8 Jul 2025)? maybe choose (Putte et al., 8 Jul 2025) because fit quality from main paper; mention G23 package without cite maybe though package paper exists. But concrete claim package paper not necessary. Could cite (Franken et al., 2023) when mentioning G23 family in dust_extinction. Fit quality still main paper maybe could cite (Putte et al., 8 Jul 2025) same section. But max once per section. So choose main paper. Maybe leave package citation aside. But developer wants relevant papers; we have already cited 2407 and 2105. Good.

Need not mention search results meta.

Let's ensure no claims not in data:

• "Python PAHFIT is a modern Python port..." yes details say. Mention context PDRs4All.
• "JWST-ready decomposition framework" exact in data.
• "reusable configurations for different object classes" yes.
• "clean separation between astrophysical model choice and instrumental resolution handling" yes. Good.
• "integrated power stays stable when FWHM changes" exact.
• "important for robust fitting of resolved JWST bands" exact.
• "many lines masked rather than fitted simultaneously" exact.
• "stellar component partly stands in for nebular continuum" exact.
• "central star-forming ring of NGC7469" yes.
• "diagnostic monotonic progression Atomic PDR → DF1 → HII → DF2 → DF3" yes.
• "The paper uses 13 fixed temperatures..." yes. Need careful with "class-A AIB" use: from conclusion. Could mention use the PDR pack for JWST PDRs / class-A AIB spectra. Fine. Maybe say "A plausible implication is ..." for analog relevance.

Let's write with numbered sections and paragraphs. Citations: Intro (Putte et al., 8 Jul 2025). Sec1 maybe no cite? Wait intro already. Section 2 cite (Zeng et al., 2021). Section 3 maybe no citation maybe since already main paper referenced in intro. But maybe okay. Section 4 cite (Kim et al., 2024). Section 5 cite (Putte et al., 8 Jul 2025). Section 6 cite (Putte et al., 8 Jul 2025). This repeats paper in different sections, okay. Up to once per section.

Need maybe one list for wavelengths. Since long enumerations, bullets acceptable. Could do concise list by complex. Since lists warranted. Fine.

Let's produce. Python PAHFIT is a spectral decomposition tool for mid-IR spectra that models the observed emission as the sum of AIB / PAH dust features, thermal dust continuum, gas emission lines, stellar continuum, and dust extinction / attenuation. It is presented as a modern Python port of the original IDL PAHFIT in the context of the PDRs4All JWST program, with the Orion Bar serving as both the main motivation and the calibration dataset for JWST-era analysis of photodissociation regions and related star-forming environments (Putte et al., 8 Jul 2025).

1. Definition and modeling scope

Python PAHFIT retains the physically motivated component-based structure of PAHFIT while expanding configurability for JWST data. Like the original PAHFIT, it fits the modeled components with least squares using Levenberg–Marquardt. Its stated purpose is to separate the contributions of AIB subcomponents, thermal dust emission, gas lines, stellar light, and dust extinction in spectra whose resolution and signal-to-noise now exceed the design assumptions of the original framework.

The decomposition target is explicitly the 3–20 $\mu$m regime in which photodissociation regions exhibit Aromatic Infrared Bands originating from small carbonaceous species such as polycyclic aromatic hydrocarbons. The paper frames Galactic PDR spectra, especially those of the Orion Bar, as local analogs for the spectra observed in extragalactic star-forming regions. This suggests that a decomposition scheme tuned on resolved Galactic templates can also serve as a starting point for unresolved galaxy spectra, provided the AIB phenomenology remains comparable.

The tool is described as a JWST-era decomposition framework with reusable configurations for different object classes. In practical terms, that framework consists of a selectable science pack, a selectable instrument pack, and a choice of continuum model, followed by extraction of physically meaningful fitted components and PAHFIT-based diagnostics.

2. Development context and departures from the original PAHFIT

The immediate developmental context is the mismatch between the original PAHFIT configuration and JWST MIRI/NIRSpec spectra. The original PAHFIT was designed for Spitzer IRS spectra at relatively low resolution, typically $R \sim 100$, whereas JWST spectra have $R \sim 1000$–3000, much higher S/N, and much more visible substructure in the AIBs (Zeng et al., 2021). The paper argues that Orion Bar templates from PDRs4All exposed three specific shortcomings of the old default setup: the original continuum model could not reproduce the steep 15–26 $\mu$m continuum of the Orion Bar, the old AIB template set was too coarse for the resolved JWST profiles, and line handling together with wavelength-dependent resolution required a more flexible framework.

A central architectural change is the introduction of “science packs” and “instrument packs.” Science packs are YAML-defined configurations listing which components to fit and with what parameters. Instrument packs define resolution curves for specific instruments and configurations. The stated purpose is to separate astrophysical model choice from instrumental resolution handling more cleanly than in the IDL implementation.

A second major change concerns the Drude parameterization of AIB features. In IDL PAHFIT, the features are specified by amplitude, central wavelength, and fractional width $\gamma_r$, with the standard Drude form

$$I^{(r)}_\nu(\lambda) = \frac{b_r \gamma_r^2}{(\lambda/\lambda_r - \lambda_r/\lambda)^2 + \gamma_r^2}.$$

Python PAHFIT instead parameterizes each feature by its total power $P_r$ and FWHM in wavelength units, with

$\gamma_r = \mathrm{FWHM}_r / \lambda_r$

and

$R_V=5.5$0

The stated consequence is that the integrated power stays stable when FWHM changes, which is described as important for robust fitting of resolved JWST bands.

A third change is improved resolution handling. Gas-line widths can be tied to the instrumental resolution curve via the instrument pack rather than adjusted more empirically. The paper presents this as a cleaner separation of spectral physics from instrument response.

3. Component model and parameterizations

The fitted model separates several classes of emission and attenuation. AIB complexes are decomposed into multiple Drude profiles, rather than treated as single bands. This is emphasized for the 3.3, 3.4, 5.2, 5.7, 6.2, 7.7, 8.6, 11.2, 12.7, and 16.4 $R_V=5.5$1m structures, which are described as multi-component complexes in JWST spectra.

The default thermal dust continuum is a sum of modified blackbodies at fixed temperatures of 35, 40, 50, 65, 90, 135, 200, and 300 K, with emissivity proportional to $R_V=5.5$2:

$R_V=5.5$3

Gas emission lines are modeled as Gaussian components. For the Orion Bar, the line density is described as sufficiently high that many lines were masked rather than fitted simultaneously. Stellar light is represented by a 5000 K blackbody. The paper states that this component is mainly relevant for extragalactic sources such as NGC7469 and is negligible overall for the Orion Bar except in the HII template, where the “stellar” component partly stands in for nebular continuum.

Dust extinction / attenuation is applied multiplicatively to the summed emission. Two geometries are discussed:

$R_V=5.5$4

The attenuation model includes a silicate-based extinction curve parameterized by $R_V=5.5$5, the optical depth at 9.7 $R_V=5.5$6m.

The paper also identifies a continuum/attenuation degeneracy. For NGC7469, fits with the standard continuum often yield $R_V=5.5$7, whereas fits with the alternate continuum yield $R_V=5.5$8–1.8. For the Orion Bar, $R_V=5.5$9 is zero in both cases. This is presented as an empirical property of the fitting outcomes rather than a resolved physical ambiguity.

4. Orion Bar calibration and the PDR pack

The main JWST-specific science pack is the “PDR pack,” a dedicated configuration for JWST PDR spectra built from Orion Bar data. The Orion Bar templates used in tuning are HII, Atomic PDR, DF1, DF2, and DF3, spanning a progression from more irradiated to more shielded conditions (Kim et al., 2024). The paper characterizes the Orion Bar as both the motivation for the Python implementation and the calibration dataset for the JWST-era pack.

The PDR pack contains 66 AIB components, compared with 26 in the classic pack. Its stated purpose is to provide more AIB subcomponents, JWST-appropriate wavelengths and FWHM values, and new broad features required to reproduce the resolved JWST structures. The paper also states that the compiled table indicates which values were taken from Atomic PDR, DF3, or averaged, and whether FWHM was fixed or allowed to vary.

The component compilation is organized by spectral complex:

• 3.3–3.6 $\mu$0m complex: 3.243, 3.280, 3.295, 3.395, 3.402, 3.425, 3.463, 3.516, 3.563 $\mu$1m.
• 5.2 and 5.7 $\mu$2m: 5.185, 5.237, 5.275, 5.441, 5.526, 5.631, 5.690, 5.754 $\mu$3m.
• 6.2 $\mu$4m: 5.880, 6.022, 6.196, 6.239, 6.341 $\mu$5m.
• 6.6–7.0 $\mu$6m approximations: 6.699, 7.030 $\mu$7m.
• 7.7 and 8.6 $\mu$8m: 7.277, 7.418, 7.552, 7.635, 7.753, 7.854, 7.896, 8.221, 8.329, 8.604 $\mu$9m.
• 10–14 $R \sim 100$0m: 10.588, 10.759, 11.008, 11.188, 11.215, 11.261, 11.340, 12.007, 11.950, 12.359, 12.611, 12.727, 12.804, 13.328, 13.561, 13.973, 14.224 $R \sim 100$1m.
• 16.4 and 17.4 $R \sim 100$2m: 15.857, 16.040, 16.400, 16.444, 16.765, 17.112, 17.382, 17.762 $R \sim 100$3m.

The significance of this compilation lies in its function as a reusable JWST-oriented template set. A plausible implication is that the pack formalizes Orion-Bar-derived substructure into a transportable parameter library for other class-A AIB emitters.

5. Alternate continuum model and fit quality

A major extension is the alternate dust continuum model introduced because the standard modified-blackbody continuum could not simultaneously match the continuum near 15 $R \sim 100$4m and the slope out to 18–26 $R \sim 100$5m. The paper states that, even with many standard modified blackbodies, the default model undercuts the 16–18 $R \sim 100$6m complex by about 10% and cannot reproduce the steep long-wavelength rise. This degrades extraction of the 16–18 $R \sim 100$7m AIB complex.

The alternate continuum replaces the $R \sim 100$8 emissivity factor with an extinction curve:

$R \sim 100$9

For the Orion Bar, the extinction curve is taken from the G23 family in the dust_extinction package with $R \sim 1000$0. The continuum grid for the alternate Orion Bar model uses 13 fixed temperatures: 45, 60, 75, 90, 105, 120, 160, 200, 240, 280, 320, 360, and 400 K. The paper further states that the alternate model introduces silicate-related structure around 10 and 20 $R \sim 1000$1m and that a single $R \sim 1000$2 K component can match the key long-wavelength slope much better than the default model (Putte et al., 8 Jul 2025).

Fit quality is reported for both the Orion Bar and the star-forming ring of NGC7469. With the PDR pack plus the alternate continuum, the Orion Bar templates are reproduced with residuals of a few percent, generally within 5%, with occasional excursions to $R \sim 1000$3, across HII, Atomic PDR, DF1, DF2, and DF3. For the seven NGC7469 star-forming ring spectra, the paper reports similar performance, with residuals typically around 10% or less. The explicit conclusion is that the PDR pack, although tuned on the Orion Bar, is a good starting point for extragalactic star-forming regions.

6. Diagnostics, photochemical sequence, and use in galaxy spectra

The paper proposes PAHFIT-based diagnostics that use fitted subcomponents to trace profile variations in the 3.3, 3.4, 5.7, 6.2, and 7.7 $R \sim 1000$4m AIBs and, by extension, the photochemical evolution of the carriers. For the 3.3 $R \sim 1000$5m band, the components are 3.280 and 3.295 $R \sim 1000$6m, and the diagnostic 3.3A is the fractional contribution of the blue component; it is stated to trace blue-wing broadening and subtle profile changes. For the 3.4 $R \sim 1000$7m band, the components are 3.395 and 3.403 $R \sim 1000$8m, and 3.4B is the fractional contribution of the red component; it traces flattening or red-side enhancement of the profile.

The 5.7 $R \sim 1000$9m complex is divided into 5.632, 5.688, and 5.752 $\mu$0m components, with diagnostics 5.7A, 5.7B, and 5.7C corresponding to the blue, middle, and red component fractions. The interpretation given is that the feature shifts from blue-dominated in more irradiated regions to red-dominated in more shielded regions. The paper stresses that the 5.7 $\mu$1m emission originates from at least two subpopulations, one more prominent in highly irradiated environments and one preferring more shielded environments.

For the 6.2 $\mu$2m complex, represented by 6.196, 6.239, and 6.341 $\mu$3m, the diagnostic is the FWHM of the summed complex. Broader profiles are interpreted as corresponding to smaller PAHs, higher excitation temperatures, or more shielded zones. For the 7.7 $\mu$4m complex, which includes multiple narrow peaks and a broad component near 7.854–7.896 $\mu$5m, the diagnostic is the broad 7.7 / total 7.7 ratio. Higher values are interpreted as indicating a larger contribution from broad emission, possibly associated with PAH clusters or very small grains, and as being stronger in shielded environments (Putte et al., 8 Jul 2025).

Across the Orion Bar templates, these diagnostics show the monotonic progression Atomic PDR $\mu$6 DF1 $\mu$7 HII $\mu$8 DF2 $\mu$9 DF3. This is interpreted as a photochemical sequence in which more irradiated regions are associated with larger, more robust PAHs and stronger blue-side or narrow components, whereas more shielded regions are associated with smaller or more processed PAHs, broader profiles, stronger red-side and broad components, and more contribution from VSG-like or broader carriers. The paper specifically associates 3.3 and 3.4 broadening with changes in PAH excitation temperature and edge structure, the 6.2 FWHM increase with smaller PAHs and higher excitation temperatures in shielded regions, the 5.7 redistribution with at least two carrier populations probably connected to PAH edge structure or hydrogen adjacency classes, and the growth of the broad 7.7 component with increasing importance of VSGs and/or PAH clusters in shielded zones.

Application to NGC7469 shows that the star-forming ring spectra are most similar to Atomic PDR, DF1, and sometimes HII, and are clearly unlike DF2 and DF3 for the better-behaved diagnostics. The interpretation given is that the mid-IR PAH emission in NGC7469 likely originates mainly from highly irradiated PDR-like regions, especially atomic-layer-like conditions. The paper also states that NGC7469’s profiles resemble the Orion Bar Atomic PDR closely enough that the Orion Bar can serve as a practical analog for unresolved galaxies.

In operational terms, the paper’s stated usage pattern is straightforward: choose a science pack, choose an instrument pack, select either the default modified-blackbody continuum or the alternate extinction-based continuum, fit the physically meaningful components, and extract diagnostic ratios or summed-band FWHM measures. The overall conclusion is that Python PAHFIT together with the PDR pack constitutes a JWST-ready decomposition framework for Galactic PDRs such as the Orion Bar, moderately obscured star-forming galaxies, and, with further refinement, a wider range of class-A AIB emitters.