{"id":"935aa5ae-256d-4e92-857b-41d720381941","arxiv_id":"2501.19397","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A spectroscopic calibration of MIRI broadband photometry recovers PAH 7.7 and 11.3 micron fluxes to within about 7% and 5% of spectral fits in four LIRGs.","lead":"Astronomers calibrated a way to measure the brightness of two key PAH emission features in galaxies using only JWST MIRI broadband images, reaching about 5 to 7 percent agreement with spectral measurements. The method lets researchers map star formation and dust properties over large areas where full spectroscopy would be too expensive.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reported ~7% and ~5% agreement is an in-sample residual from calibrating g_cont (and the Ssil,phot cutoff) on the same 267 spectra, so the claimed accuracy is not yet an out-of-sample result; transferability to other galaxy populations is the central untested assumption.","rationale":"The reader's weakest-assumption analysis and my stress-test converge on the same issue: the 7% and 5% accuracy figures are computed on the calibration sample itself. I checked the internal consistency of the prescription: Eqs. 5 and 6 follow from Eq. 4 and the constants in Table 1, and the numerical prefactors are consistent with the stated w_b', c_PAH, c_wing products, so I do not see an arithmetic or logical error in the derivation. The method is honest and clearly describes how g_cont, Eq. 8, and the Ssil,phot cutoff were derived. However, the headline claim is about predictive accuracy for new targets, and that claim is supported only by in-sample residuals. The calibration sample is a specific population (LIRGs) with a limited range of physical conditions; the paper provides no hold-out test, no cross-validation, and no comparison with real MIRI imaging (rather than synthetic photometry from spectra). This is not an internal inconsistency, but it is a correctness risk for the general applicability of the prescription. Because the reader already assigned CONDITIONAL, my recommendation is UNCHANGED: the paper is valuable but should be conditioned on an external or cross-validated demonstration that the calibrated constants transfer. This is a concrete, feasible check using the authors' own existing data (LOOCV) or a modest amount of new archival data, and it would settle whether the in-sample accuracy generalizes.","tokens_in":17545,"tokens_out":5605,"duration_ms":56262,"concrete_test":"Perform leave-one-galaxy-out cross-validation: recalibrate g_cont for each filter combination, the Eq. 8 coefficients, and the Ssil,phot cutoff using any three of the four GOALS galaxies, then apply the resulting prescription to the held-out galaxy and compute median percent differences vs. CAFE fits. If the held-out median absolute percent difference exceeds ~10% for PAH 7.7 µm or ~7% for PAH 11.3 µm, or the held-out IQR exceeds the in-sample IQR by more than a factor of 1.5, the reported accuracy is not representative of out-of-sample performance. A complementary check is to apply the published prescription to MIRI imaging plus MRS spectra of a non-GOALS star-forming galaxy (e.g., a PHANGS-JWST target) and compare the photometric PAH fluxes to spectral decomposition.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Abstract; §5) is that the prescription yields PAH 7.7 µm and 11.3 µm fluxes agreeing with spectral decomposition to ~7% and ~5%. Those numbers come from the same spectra used to calibrate the method. In §4.2, g_cont is defined as the median ratio between the CAFE continuum and the power-law pseudo-continuum for the sample; applying this median necessarily centers the in-sample percent-difference distribution near zero. Thus the quoted 'Cal.' columns in Table 2 are residual scatter after removing the sample-specific systematic offset, not an independent measure of predictive accuracy. The PAH 11.3 µm adjustment in §5 is further tuned in-sample: Eq. 8 is derived from a fit to the same spectra, and the Ssil,phot < -0.6 switch was chosen by testing cutoffs between -2.5 and 0 and selecting the best performer on this sample. The calibration sample is four GOALS LIRGs (NGC 3256, NGC 7469, VV 114, II Zw 96), spanning SFR densities 0.1-3.1 Msun/yr/kpc^2 and Ssil between -2.77 and 0.23. Whether the power-law offset g_cont and the silicate relation are invariant across this parameter space, or beyond it (normal star-forming galaxies, lower metallicity, different dust temperatures), is assumed but not demonstrated. The redshift test in §5 shifts three spectra in wavelength but does not change the underlying continuum/PAH shape or environment, so it does not test environmental transferability. Although the constants c_PAH, c_wing, and w_b' vary by only ~1-2% across the sample and are physically motivated, the same cannot be said of g_cont, which ranges from 0.49 to 1.56 across filter combinations in Table 2. If the true continuum-to-pseudo-continuum offset varies with galaxy type or local conditions, the 5-7% claim will not generalize. This is the load-bearing assumption on which the central claim rests.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a photometric prescription for extracting PAH 7.7 and 11.3 micron fluxes from JWST MIRI broadband imaging. Using 267 MIRI/MRS spectra from four GOALS LIRGs, the authors derive synthetic MIRI filter photometry, estimate a power-law pseudo-continuum anchored in continuum-dominated bands, and apply empirically calibrated constants (w_b', c_PAH, c_wing, and a continuum correction g_cont) to convert photometric excesses into PAH complex fluxes. They compare the photometric estimates with fluxes from CAFE spectral decomposition and report typical agreements of about 7% for PAH 7.7 and about 5% for PAH 11.3, after applying a silicate-strength-dependent correction for the 11.3 micron complex.","tokens_in":18081,"tokens_out":4335,"duration_ms":44940,"significance":"If the claimed accuracy holds beyond the calibration sample, the prescription would be practically valuable: MIRI imaging offers a much larger field of view and higher sensitivity to faint extended emission than MRS IFU maps, so the method would enable high-resolution PAH flux maps across entire galaxies. The paper is transparent about its methods: the constants are physically motivated, the synthetic-photometry procedure is clearly specified, and the dependence of performance on six filter combinations is tabulated. The explicit silicate correction and the redshift-sensitivity test are useful additions. The main limitation is that the headline accuracy figures are measured on the same spectra used to calibrate g_cont, the Eq. 8 coefficients, and the S_sil cutoff, so they are in-sample residuals rather than out-of-sample predictions; the transferability of the calibration to other galaxy populations is asserted but not demonstrated.","major_comments":[{"comment":"The reported ~7% and ~5% accuracies are in-sample residuals, not independent predictive errors. The correction g_cont in Table 2 is defined as the median ratio of the CAFE continuum to the power-law pseudo-continuum over the same 267 spectra that are then used to compute the 'Cal.' columns and Figure 5. Applying a sample median as a correction necessarily centers the percent-difference distribution near zero and reduces the IQR by construction. To support the abstract claim that the prescription 'yields predicted flux densities' with this accuracy, the authors should provide a leave-one-galaxy-out or split-half cross-validation, or an application to independent MIRI imaging plus spectroscopy, and report the resulting residuals and the stability of the fitted constants.","section":"§4.2, Table 2, Figure 5"},{"comment":"The improved PAH 11.3 prescription is tuned on the same calibration sample: the coefficients in Eq. 8 come from a fit to the 267 spectra, and the S_sil,phot < −0.6 switch is chosen by testing cutoffs between −2.5 and 0 and selecting the best performer on this sample. Therefore the quoted ~5% accuracy with the combined Eq. 6/Eq. 8 procedure is a selected in-sample result, and the extent to which the threshold and coefficients generalize is unknown. A cross-validation scheme that treats the cutoff selection as part of the fitting procedure is needed before the accuracy can be claimed as predictive.","section":"§5, Eqs. 7–8, Figure 6"},{"comment":"The transferability of g_cont and the silicate correction to galaxies outside the calibration sample is the central untested assumption. The sample consists of four GOALS LIRGs spanning SFR densities 0.1–3.1 Msun/yr/kpc^2 and S_sil from −2.77 to 0.23, but normal star-forming galaxies, low-metallicity systems, and regions with different dust temperatures or PAH fractions may have a different offset between the power-law pseudo-continuum and the true continuum. The redshift test in §5 shifts only three spectra in wavelength and does not change the underlying continuum shape or environment, so it does not test environmental transferability. I would like to see a concrete test, for example applying the prescription to archival MIRI imaging plus MRS data of nearby non-LIRG galaxies, or an explicit analysis of residual scatter against S_sil, SFR density, and metallicity within the present sample.","section":"§4.2, §5, Table 2"}],"minor_comments":[{"comment":"The filter label 'F0560W' appears in several rows and in the table header; it should be 'F560W'.","section":"Table 2"},{"comment":"The sentence 'Thus, we artifically set gcont = 1 for these combinbations in Table 2' contains typos ('artifically', 'combinbations') and should read 'artificially' and 'combinations'; nearby text also has 'discusssed'.","section":"§4.2"},{"comment":"The sentence reporting improved accuracy for low-to-moderate absorption gives the format as '−3.6 (−1.6, −5.6)%'; if these are median followed by quartiles, the quartile order appears reversed, and the convention should be clarified.","section":"§5"},{"comment":"Figure 4's caption states g_cont = 0.86 and 0.88 for the two F1000W-anchored PAH 11.3 combinations, while Table 2 lists g_cont = 1 for those combinations with a dagger. The relationship between the tabulated values and the values used in Figures 4 and 5 should be explained consistently.","section":"§4.2, Fig. 4 caption"},{"comment":"The artificial-redshift test is applied to only three spectra; the text should state which spectra were used and how representative they are of the sample, since the resulting redshift limits may depend on the chosen continuum and silicate strengths.","section":"§5"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the manuscript is a useful calibration paper for PAH photometry with MIRI, and the in-sample performance is encouraging. The main issue is that the headline accuracies (7% and 5%) are presented as predictive while being measured on the calibration sample; this should be fixed by reframing the claims as in-sample precision and providing cross-validation or an independent test. If the authors cannot obtain independent validation data, they should soften the abstract and conclusions accordingly. The paper is within the scope of an astrophysics journal and does not have any apparent novelty-disclosure problems."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a genuinely useful calibration paper. The new work relative to Chown et al. is the extension to extragalactic LIRGs, the systematic comparison of six MIRI filter combinations, and the silicate-adjusted 11.3 micron prescription. The constants cPAH, cwing, and wb' are physically motivated and vary little across the sample, which is reassuring. The paper is clearly written, transparent about the sample and cuts, and the equations are straightforward to apply. The redshift test is a nice robustness check, even if limited to three spectra.\n\nThe soft spot is exactly where the reader and stress-test put it: the headline accuracy numbers are in-sample. gcont is defined as the median ratio between the CAFE continuum and the power-law pseudo-continuum over the same 267 spectra used to report the ~7% and ~5% agreement. Applying that median necessarily centers the percent-difference distribution near zero, so Table 2 and Figure 6 are showing residual scatter after absorbing a sample-specific offset, not out-of-sample predictive accuracy. The Ssil,phot < -0.6 switch was also chosen by scanning cutoffs on this sample. The calibration sample is four GOALS LIRGs, spanning SFR density 0.1-3.1 Msun/yr/kpc^2 and Ssil from -2.77 to 0.23, but transferability to normal star-forming galaxies, lower metallicities, or very different dust temperatures is assumed, not demonstrated. The redshift test does not change the underlying spectral shape or environment.\n\nThat said, the central argument holds up as a calibration paper. The authors do not overclaim out-of-sample generality; they present a prescription calibrated on a specific sample. The issue is that the abstract and §5 phrase the accuracy as if it were the precision an end user should expect for arbitrary new targets. That is probably close to true in similar LIRG-like conditions, but it is not yet established.\n\nMy recommendation: send it out for serious review. The method is useful and the community will want it. Ask the authors to label the accuracy as in-sample, and to add a leave-one-galaxy-out cross-validation or an application to real MIRI images rather than synthetic photometry. Even without that, the paper merits publication; with a modest validation, it becomes a standard reference.","headline":"Useful, honest calibration paper; the headline ~7%/~5% accuracy is in-sample residual scatter, so treat it as calibration performance until an independent validation appears.","tokens_in":18763,"tokens_out":1702,"would_cite":true,"duration_ms":19589,"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 introduces a photometry-only prescription that recovers PAH 7.7 and 11.3 micron flux from JWST MIRI broadband images, with typical agreement of ~7% and ~5% with spectral decomposition.","keywords":["PAH flux","JWST MIRI imaging","synthetic photometry","spectral decomposition","luminous infrared galaxies","silicate absorption","star formation tracers","mid-infrared continuum"],"falsifier":"Compare the prescription's PAH 7.7 and 11.3 micron fluxes against MIRI/MRS spectral fits for nearby galaxies outside the calibration set, especially low-metallicity dwarfs and AGN-dominated nuclei; a median absolute percent difference above about 7% or 5%, or a trend in the residuals with silicate strength or spectral slope, would show the calibration is sample-specific.","tokens_in":17341,"feed_emoji":"🌌","tokens_out":10804,"duration_ms":92544,"temperature":0.7,"pith_summary":"Polycyclic aromatic hydrocarbons (PAHs) trace the radiation field and star formation in galaxies, but mapping their 7.7 and 11.3 micron features normally requires slow integral-field spectroscopy. This paper argues that the same fluxes can be recovered from JWST MIRI broadband images alone, with accuracy comparable to the spectral fits themselves: within about 7% for the 7.7 micron complex and 5% for the 11.3 micron complex. The recipe anchors a power-law pseudo-continuum between two continuum-dominated MIRI filters, subtracts it from the PAH-dominated band, and applies constants calibrated on 267 MIRI/MRS spectra from four luminous infrared galaxies. If the calibration transfers, PAH flux maps over fields of view more than 100 times larger than the MIRI integral-field unit become routine, enabling resolved studies of PAH grain size and ionization across whole galaxies.","feed_headline":"MIRI images alone recover PAH flux to within 5-7 percent","feed_subtitle":"A four-filter recipe reproduces spectral-fit PAH 7.7 and 11.3 micron fluxes on 267 spectra","key_machinery":"The load-bearing object is the spectroscopically calibrated photometric identity $F_{\\mathrm{PAH},b'} = 10^{-26}\\nu_{b'} w_{b'} c_{\\mathrm{PAH}} c_{\\mathrm{wing}} (f_{b'} - g_{\\mathrm{cont}} f_{\\mathrm{blue}}^{(1-\\alpha)} f_{\\mathrm{red}}^{\\alpha})$, where the power-law pseudo-continuum $f_{\\mathrm{blue}}^{(1-\\alpha)} f_{\\mathrm{red}}^{\\alpha}$ estimates the local mid-infrared continuum from two continuum-dominated MIRI bands, $g_{\\mathrm{cont}}$ is the median ratio of the spectrally fitted continuum to that pseudo-continuum in the PAH band, and the three dimensionless constants convert filter flux density into integrated PAH-complex flux. The power-law construction is what allows broadband images to stand in for spectroscopy; the $g_{\\mathrm{cont}}$ calibration removes the systematic offset between a featureless power law and the true dust continuum; and for PAH 11.3 the additional silicate-strength term handles the 9.7 micron absorption feature that would otherwise bias the blue anchor.","core_discovery":"On the paper's own terms, the central discovery is that the observed integrated fluxes of the 7.7 and 11.3 micron PAH complexes can be recovered from JWST MIRI broadband imaging with typical agreement of about 7% and 5%, respectively, compared with values from detailed spectral decomposition, matching the roughly 5% uncertainty of the spectroscopic fits themselves. The recommended filter combinations are F560W, F770W, and either F1500W or F2100W for the 7.7 micron complex, and F560W, F1000W, F1130W, and F1500W for the 11.3 micron complex. The method models the underlying continuum as a power law anchored in continuum-dominated bands, subtracts a median-calibrated version of that power law from the PAH-dominated band, and applies constants accounting for the filter width, the fraction of the PAH complex inside the band, and power lost outside the filter through the broad wings of the PAH profiles. For strongly silicate-absorbed regions the prescription adds a photometric estimate of silicate strength and replaces the blue continuum anchor with an absorption-corrected 10 micron flux.","pith_inferences":["If the continuum-correction transferability assumption holds, the prescription can be tested immediately on archival MIRI imaging of nearby galaxies that already have MIRI/MRS spectra, without any new observations.","Because the calibration set covers only four luminous infrared galaxies, the claimed 5-7% accuracy may not extend to low-metallicity dwarfs or strongly AGN-dominated regions; recomputing the constants on such spectra would quantify any drift.","The same power-law-and-correction structure could plausibly be adapted to other PAH-dominated MIRI bands, such as F1280W for the 12.6 micron feature, though the silicate absorption treatment would need to be re-derived there.","The silicate-strength cutoff and the exponential relation used in the absorption correction are tuned to this sample; applying the formula to more heavily obscured systems would test whether those relations are universal or specific to the calibration galaxies."],"forward_implications":["MIRI imaging alone can produce PAH 7.7 and 11.3 micron flux maps over fields of view more than 100 times larger than the MIRI integral-field unit, at the same spatial resolution, without spending the observing time needed for spectral mapping.","The recommended filter combinations recover PAH flux at roughly the same accuracy as the spectral fits themselves, so resolved PAH ratios can be mapped across whole galaxies to trace grain size and ionization state.","For local-universe targets the method degrades predictably with redshift: PAH 7.7 flux is under-recovered by about 2% at z=0.01, 7% at z=0.03, and 8-16% at z=0.06, while PAH 11.3 stays within about 2.5% only until z is near 0.015.","Combining these maps with PAH 3.3 micron flux from NIRCam photometry yields spatially resolved probes of PAH grain size and ionization over large areas of nearby galaxies."],"supporting_citations":[{"why":"Demonstrated that MIRI photometry can recover PAH flux at parsec scales, providing the proof of concept this paper extends to galaxy-scale regions.","marker":"Chown et al. (2024b)"},{"why":"Introduced the two-band power-law pseudo-continuum that the paper's Eq. (1) uses to estimate the mid-infrared continuum.","marker":"Marble et al. (2010)"},{"why":"Defines the multi-Drude PAH complexes at 7.7 and 11.3 micron and the spectral decomposition framework whose fluxes are the reference values.","marker":"Smith et al. (2007)"},{"why":"Underlies the continuum-and-feature extraction method used to fit the spectra and measure the comparison PAH fluxes.","marker":"Marshall et al. (2007)"},{"why":"Supplies the 6.2 micron equivalent-width criterion and the silicate strength S_sil used to cut the sample and calibrate absorption corrections.","marker":"Spoon et al. (2007)"},{"why":"Characterizes the 9.7 micron silicate absorption feature whose shape is the reason the F1000W band is needed for PAH 11.3.","marker":"Kemper et al. (2004)"},{"why":"Provides the synthetic photometry conventions and pivot wavelengths used to convert spectra into filter flux densities.","marker":"Gordon et al. (2022)"}],"fun_headline_variants":["MIRI imaging alone recovers PAH flux within 5-7% of spectra","A filter recipe pulls PAH 7.7 and 11.3 flux from MIRI images","Imaging-only PAH flux: within 7% (7.7 µm) and 5% (11.3 µm) of spectral fits","Spectrally calibrated MIRI images estimate PAH flux to 5-7%","MIRI imaging: a calibrated shortcut to PAH flux within 5-7%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole recipe depends on the assumption that the median correction between a simple power law and the true mid-infrared spectrum, measured in four bright merging galaxies, works for any other galaxy; if that offset is different elsewhere, the claimed 5-7 percent accuracy will not transfer.","fun_headline_variants_meta":{"raw":{"variants":["MIRI imaging alone recovers PAH flux within 5-7% of spectra","A filter recipe pulls PAH 7.7 and 11.3 flux from MIRI images","Imaging-only PAH flux: within 7% (7.7 µm) and 5% (11.3 µm) of spectral fits","Spectrally calibrated MIRI images estimate PAH flux to 5-7%","MIRI imaging: a calibrated shortcut to PAH flux within 5-7%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00059,"raw_usage":{"total_tokens":2816,"prompt_tokens":1039,"completion_tokens":1777,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":655,"completion_tokens_details":{"reasoning_tokens":1647}},"tokens_in":655,"tokens_out":1777,"duration_ms":11147,"temperature":1.0,"reasoning_tokens":1647,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T20:14:08.351607+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the prescription's PAH 7.7 and 11.3 micron fluxes against MIRI/MRS spectral fits for nearby galaxies outside the calibration set, especially low-metallicity dwarfs and AGN-dominated nuclei; a median absolute percent difference above about 7% or 5%, or a trend in the residuals with silicate strength or spectral slope, would show the calibration is sample-specific.","supporting_citations":[],"review_version":1}