{"id":"365ebf78-dfb8-4b10-a860-fb23c7627e83","arxiv_id":"2412.11627","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Combined JWST/MIRI eclipses at 12.8 and 15 µm give depths of 452±86 and 775±90 ppm, leaving an airless ultramafic surface and a hazy CO2 atmosphere as equally viable explanations.","lead":"TRAPPIST-1 b's thermal emission measured in two mid-infrared bands is best explained by either a bare rock with a fresh ultramafic surface or a thick hazy CO2 atmosphere, and two bands alone cannot tell them apart. The new 12.8 µm eclipse depth, combined with the previous 15 µm data, shows the planet is more complex than a simple zero-albedo bare rock.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The two-scenario interpretation rests on fits that use the NIRISS/SOSS transmission spectrum as a constraint, even though the paper itself identifies that spectrum as strongly stellar-contaminated; refitting with only the MIRI eclipse depths could alter or dissolve the hazy-CO2 scenario.","rationale":"The paper is careful in many ways: four independent reductions, four analyses, honest discussion of the haze-formation caveats, and a clear statement that the hazy atmosphere is less likely than the bare-surface scenario. I therefore do not want to manufacture a fatal flaw. The single most load-bearing issue is the undisclosed use of a known-contaminated transmission spectrum in the model fits that produce the two-scenario interpretation. The paper itself provides the ammunition for this concern by arguing that TLS contamination makes transmission spectra unreliable for TRAPPIST-1 b, then using such a spectrum as a joint constraint in the retrievals. The haze-formation feasibility is a genuine caveat, but it is already disclosed in the text and the authors explicitly rank the airless scenario as more likely, so it is less damaging to the central claim as framed. The contaminated-spectrum issue is not disclosed in the retrieval sections and could change which scenario is preferred. Still, because the eclipse-depth measurement is robust and the authors already present the interpretation conditionally, the correct verdict remains CONDITIONAL; my stress-test does not change the reader's verdict.","tokens_in":28058,"tokens_out":6968,"duration_ms":66769,"concrete_test":"Rerun the Bond-albedo fit and the hazy-CO2 retrieval using only the two MIRI eclipse depths (and the measured stellar fluxes), omitting the NIRISS/SOSS transmission spectrum entirely. If the hazy-CO2 model still fits both eclipse depths within 1σ and the recovered fhaze and albedo remain within 2σ of the published values, the concern does not land. If the fit degrades substantially or the parameters shift by more than 2σ, the two-scenario conclusion depends on the contaminated transmission spectrum and should be reworded to separate the robust eclipse-depth result from the model-dependent scenarios.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central eclipse depths (452 ± 86 ppm and 775 ± 90 ppm) are supported by four independent reductions and multiple fitting methods, so I do not question the measurement itself. The load-bearing weakness is in the interpretation stage. In Methods (sections 'Bare surfaces' and 'Atmospheric models'), both the Bond-albedo fit and the hazy-CO2 retrieval jointly fit the MIRI eclipse depths together with the NIRISS/SOSS transmission spectrum of ref. 51. The paper's own Introduction (Section 1.1) states that the NIRISS transmission spectrum of TRAPPIST-1 b shows strong stellar contamination via the transit light source effect, and argues that emission photometry is the preferred avenue for this system. Using that contaminated spectrum as a constraint can bias the retrieved radius, Bond albedo, haze mass fraction, and band gap, and therefore bias the comparison between the airless-surface and hazy-atmosphere scenarios. The reduced-chi-square surface comparison in Supplementary Table 1 is independent of the transmission spectrum, but the quoted Ab = 0.19 ± 0.08 and the atmospheric retrieval are not. The erroneous citation of the optEC(s) haze optical-properties model (ref. 53 is a scattered-moonlight paper) adds further uncertainty to the atmospheric fit, though the ARCiS/ATMO comparison suggests the inversion itself is robust. If the hazy-CO2 model only fits because it absorbs the stellar contamination into fhaze/Eg, the 'two main scenarios' claim is overstated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents five new JWST/MIRI secondary-eclipse observations of TRAPPIST-1 b at 12.8 µm, combined with a re-analysis of five previously observed eclipses at 15 µm. A global fit of all ten eclipses yields eclipse depths of 452 ± 86 ppm at 12.8 µm and 775 ± 90 ppm at 15 µm, corresponding to brightness temperatures of 424 ± 28 K and 478 ± 27 K. The authors then explore two model interpretations: an airless ultramafic surface with a fitted Bond albedo of 0.19 ± 0.08, and a thick, pure-CO2 atmosphere with photochemical hazes producing a thermal inversion. They argue that both scenarios fit the data and that future phase-curve observations can distinguish them.","tokens_in":28390,"tokens_out":6430,"duration_ms":58889,"significance":"The central measurement is a genuine advance: four independent reductions and four distinct analysis methods give mutually consistent eclipse depths, and the data and source products are archived. The 12.8 µm measurement in particular provides a key constraint inside the CO2 band and reveals a possible tension with the previously favored null-albedo bare-rock interpretation. If the two-scenario conclusion holds, the paper will be an important benchmark for broadband emission studies of rocky exoplanets. However, the interpretation stage is less secure: both the Bond-albedo fit and the hazy-CO2 retrieval use the NIRISS/SOSS transmission spectrum as a joint constraint, even though the paper itself argues that this spectrum is strongly contaminated by the transit light source effect. The citation for the haze optical properties is also incorrect. The measurement should stand, but the interpretive claims need strengthening before publication.","major_comments":[{"comment":"The hazy-CO2 retrieval (fhaze = 4.5×10−4, Eg, Rp) uses the NIRISS/SOSS transmission spectrum of ref. 51 as a constraint, even though Section 1.1 argues that this spectrum is strongly contaminated by the transit light source effect. Because the transmission spectrum is used to fit the same model parameters that then produce the claimed good fit to the eclipse depths, the atmospheric scenario is not an independent MIRI-only result. Please re-run the retrieval using only the two MIRI eclipse depths (or a TLS-marginalized treatment of the NIRISS data) and report whether the hazy-CO2 solution survives; if it does not, the 'two main scenarios' conclusion must be revised to a single preferred scenario.","section":"Methods, '1D atmosphere with full heat redistribution'; Section 1.1"},{"comment":"The quoted Bond albedo Ab = 0.19 ± 0.08 is derived from a joint fit to both the MIRI eclipse depths and the NIRISS transmission spectrum. This is circular with respect to the claim that the airless model fits the MIRI data well: the albedo is adjusted to match the eclipse depths rather than predicted from them. The only MIRI-only model comparison in the paper is the reduced-χ² table for the fixed surface-composition models (Supplementary Table 1), which does not constrain Ab. Please separate the MIRI-only albedo constraint from the joint-fit value and quantify how the TLS-contaminated transmission spectrum affects Ab.","section":"Methods, 'Bare surfaces'"},{"comment":"Reference 53 is Jones et al. (2013), 'An advanced scattered moonlight model for Cerro Paranal,' which does not contain the optEC(s) hydrocarbon haze optical properties used for the haze opacity. The correct source for these optical constants must be identified and cited, or the constants provided as supplementary data; without this, the thermal inversion calculation is not reproducible.","section":"Methods, 'Atmospheric models'"},{"comment":"The statement that the hazy-CO2 model 'can fit the measurements very well' is not supported by a quantitative goodness-of-fit or model-comparison statistic for the atmospheric model. Please provide a chi-squared, Bayesian evidence, or equivalent for the atmospheric scenario against the MIRI data, alongside the existing bare-surface χ²_r values.","section":"Discussion, third paragraph"}],"minor_comments":[{"comment":"Equation (1) integrates cosθ cosφ dθ dφ with θ defined as longitude and φ as latitude; this appears to contain an extra geometric factor or a missing area element. Please check the projection and area element and correct the equation or the coordinate definitions, as it affects the computed blackbody flux.","section":"Methods, 'Bare surfaces', Eq. (1)"},{"comment":"The caption says the measurements are 'compared to realistic emission models for bare surface models,' but the right-hand panel of the figure contains atmospheric models; please make the caption consistent with the two panels.","section":"Figure 2 caption"},{"comment":"The text contains broken variable names such as 'Nf rames' and 'aper-ima'; these should be formatted correctly for readability.","section":"Methods, 'Data Reduction: POL'"},{"comment":"There are several typographical errors, e.g., 'fist and last groups' and 'four the four others'; please proofread the Methods section.","section":"Methods, 'Data Reduction: ED'"}],"recommendation":"major_revision","confidential_remarks":"The eclipse-depth measurement is robust and well documented; the main risk to the paper's central claim lies in the interpretation stage, where the hazy-CO2 and Bond-albedo results depend on the NIRISS transmission spectrum that the authors themselves flag as strongly contaminated. The incorrect citation of the haze optical-properties source is easily fixed but should be corrected before acceptance. I would regard a MIRI-only retrieval as a required revision, not merely a suggestion."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Elsa and colleagues have done the important part: five new 12.8 micron eclipses, combined with the five 15 micron ones, four independent reductions, four fitting methods, consistent depths. The final values, 452±86 and 775±90 ppm, are the kind of measurement you can build on. The 12.8 micron depth coming in below the zero-albedo blackbody prediction is a real result, and the 15 micron depth agrees with Greene et al. So the central measurement holds.\n\nThe surface comparison is also fair: the ultramafic surface model wins the reduced chi-square comparison among the bare-surface models, and that comparison does not use the transmission spectrum. So the 'fresh ultramafic surface' scenario stands on its own.\n\nThe soft spot is the atmospheric interpretation. The Bond albedo fit and the hazy-CO2 retrieval both use the NIRISS/SOSS transmission spectrum as a constraint, and the paper itself says that spectrum is strongly stellar-contaminated. Fitting a contaminated transmission spectrum alongside two eclipse depths can push fhaze and Eg to soak up the contamination. The stress test is right: the quoted Ab=0.19 and the hazy-scenario parameters are not independent of that contaminated data. There is also an outright citation error: ref 53 is a scattered-moonlight paper, not the optEC(s) haze model. The ARCiS vs ATMO comparison suggests the thermal inversion itself is robust, so the hazy scenario is not a phantom, but the specific fit values and the 'fits very well' claim are conditioned on that contaminated input.\n\nThe paper is honest about the circularity and the speculative haze chemistry, and it flags the H2S caveat itself. I do not think the two-scenario framing is malicious, but it is overstated: the airless ultramafic surface is the better-supported scenario, and the hazy one needs a re-fit without the NIRISS spectrum before it can be called a co-equal explanation.\n\nWho is this for: anyone interpreting JWST emission photometry of rocky planets, and anyone doing retrievals on M-dwarf targets where the transmission spectrum is contaminated. It deserves a serious referee. I would send it out, with a request to redo the atmospheric retrieval without the NIRISS spectrum, correct the optEC(s) citation, and soften the 'two main scenarios' claim accordingly.","headline":"A robust new eclipse-depth measurement for TRAPPIST-1 b, with an interpretation that is more fragile than the paper's two-scenario framing suggests.","tokens_in":29141,"tokens_out":2312,"would_cite":true,"duration_ms":22556,"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":"Ten JWST eclipses of TRAPPIST-1 b yield two brightness temperatures that fit both a bare ultramafic rock and a hazy CO2 atmosphere, with a phase curve predicted to break the tie.","keywords":["secondary eclipse photometry","TRAPPIST-1 b","JWST MIRI","thermal emission","bare rocky exoplanet","CO2 atmosphere","photochemical haze","thermal inversion"],"falsifier":"A single 15 µm phase curve (the planned GO 3077 observations) would settle the matter: the airless ultramafic model predicts an almost vanishing nightside flux and a symmetric eclipse (night-to-day flux near zero), whereas the 10-bar hazy CO2 model predicts a night-to-day ratio of 0.85 and a phase-curve offset; if the measured nightside flux is near zero, the hazy-atmosphere scenario is ruled out.","tokens_in":27875,"feed_emoji":"🔭","tokens_out":7221,"duration_ms":61206,"temperature":0.7,"pith_summary":"TRAPPIST-1 b, an Earth-sized rocky planet around an ultra-cool dwarf, was observed in ten secondary eclipses with JWST/MIRI: five in a 12.8 µm filter and five in a 15 µm filter. Combining all ten, the paper measures planet-to-star flux ratios of 452 ± 86 ppm at 12.8 µm and 775 ± 90 ppm at 15 µm, corresponding to brightness temperatures of 424 ± 28 K and 478 ± 27 K. The two bands were chosen to straddle the CO2 absorption feature, but the data do not select a single story: they are consistent either with an airless planet whose surface is fresh, unweathered ultramafic rock, or with a thick, pure-CO2 atmosphere containing photochemical hazes that create an upper-atmosphere temperature inversion and make CO2 emit rather than absorb. The paper shows that two very different worlds can produce the same two broadband measurements, and argues that a forthcoming phase curve should distinguish them.","feed_headline":"TRAPPIST-1 b fits both a bare rock and a hazy CO2 sky","feed_subtitle":"Ten MIRI eclipses give two brightness temperatures, and both an airless ultramafic surface and a haze-warmed atmosphere match.","key_machinery":"The argument runs on the spectral contrast between the two MIRI filters, F1280W (11.6–14.2 µm) and F1500W (13.5–16.6 µm), centered inside and outside the 15 µm CO2 band. On one side, a grid of radiative-convective surface models (basaltic, ultramafic, feldspathic, metal-rich, Fe-oxidized, granitoid) predicts band-integrated eclipse depths; the ultramafic composition (60% olivine, 40% enstatite) matches the observed ratio, while space-weathering expectations make a young surface plausible. On the other, a 1D radiative-transfer retrieval that mixes hydrocarbon haze particles (50 nm particles with optEC(s) optical properties) into a pure-CO2 atmosphere produces a temperature inversion that flips the CO2 feature from absorption to emission, and the same haze abundance also fits the NIRISS transmission spectrum. The proposed discriminator is heat redistribution: the airless model has essentially no nightside flux, whereas the hazy atmosphere redistributes a large fraction of absorbed stellar energy, so the predicted 15 µm phase curve differs sharply between the two cases.","core_discovery":"The central claim is that the 12.8 µm and 15 µm eclipse depths of TRAPPIST-1 b, analyzed together for the first time, are each individually consistent with more than one physical picture. The authors establish that the brightness temperature at 12.8 µm (424 ± 28 K) is 2.1σ lower than expected for a dark bare rock, implying a Bond albedo of 0.19 ± 0.08. They then show that a bare, geologically fresh ultramafic surface fits the two eclipse depths, and that a hazy CO2-dominated atmosphere with a thermal inversion also fits, producing CO2 in emission at 15 µm. The paper's conclusion is that this two-scenario degeneracy cannot be resolved with only two broadband points, and that the heat-redistribution signal expected from phase-curve observations will tell an airless world apart from a hazy, heat-redistributing atmosphere.","pith_inferences":["The authors stop short of saying so, but the same degeneracy should affect other emission-photometry surveys of rocky M-dwarf planets: any two broadband points can be matched by some surface composition or some haze-affected atmosphere, so phase curves or resolved spectra will be needed before claiming an airless or atmospheric detection.","If the fresh ultramafic surface is the right answer, TRAPPIST-1 b would join Io as a volcanically resurfaced world; a testable corollary is that eclipse depth at 12.8 µm could vary on resurfacing timescales, which is exactly the currently insignificant variability hinted at in the five individual eclipses.","The haze path could be checked before more telescope time is spent: laboratory or photochemical-model experiments scanning H2S abundance, CO2/O2 ratio, and TRAPPIST-1-like XUV flux would show whether the required 1% H2S and the resulting haze column are realistic, putting a prior on which of the two scenarios to prefer."],"forward_implications":["A single pair of broadband eclipse depths cannot by itself determine whether TRAPPIST-1 b is airless or has a thick atmosphere; the paper predicts that the upcoming 15 µm phase curve (program GO 3077) will separate the two.","If the airless interpretation is correct, the surface must be relatively reflective (Bond albedo 0.19 ± 0.08) and geologically fresh ultramafic rock, implying recent volcanic or tectonic resurfacing on a tidally and induction-heated world.","If the hazy atmosphere interpretation is correct, hazes produce a strong thermal inversion that makes the 15 µm CO2 band appear in emission, and the needed haze formation pathway depends on volcanic H2S at roughly the 1% level.","The 12.8 µm brightness temperature is 2.1σ lower than the null-albedo bare-rock prediction, so neither a dark blackbody nor a simple greenhouse atmosphere matches; both scenarios require a departure from the simplest expectations.","The joint fit of all ten eclipses confirms the earlier 15 µm detection at the 1σ level while tightening the eclipse depth, and statistical tests find no significant eclipse-depth variability at 12.8 µm."],"supporting_citations":[{"why":"Supplies the five 15 µm eclipses and the earlier eclipse-depth detection that this work reanalyzes and extends with the 12.8 µm data.","marker":"ref. 12"},{"why":"Provides the bare-surface emission models (basaltic, ultramafic, granitoid, etc.) that define the airless-scenario predictions.","marker":"ref. 25"},{"why":"Laboratory demonstration that hydrocarbon hazes can form in hot CO2-rich atmospheres without methane if H2S is present; load-bearing for the hazy-atmosphere scenario.","marker":"ref. 35"},{"why":"Provides the 1D pure-CO2 atmosphere model setup that the haze retrieval modifies to produce a thermal inversion.","marker":"ref. 49"},{"why":"Supplies the NIRISS transmission spectrum used as an additional constraint in both the albedo and haze retrievals.","marker":"ref. 51"},{"why":"Supplies system parameters, ephemerides, transit-timing masses, and priors used in the joint eclipse fits.","marker":"ref. 9"},{"why":"Mega-MUSCLES spectral energy distribution used to argue that TRAPPIST-1 b receives Titan-like UV flux, supporting haze photochemistry.","marker":"ref. 32"},{"why":"3D diffusion model used to simulate phase curves and predict the pressure-dependent heat redistribution that will discriminate the scenarios.","marker":"ref. 56"}],"fun_headline_variants":["Two wavelengths, two faces: bare rock or hazy CO2","TRAPPIST-1 b's dual readings fit airless and hazy models","Bare rock or hazy CO2? Both fit TRAPPIST-1 b","New eclipses double TRAPPIST-1 b's possible worlds","Hazy CO2 atmosphere or bare rock? TRAPPIST-1 b keeps both"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The hazy CO2 scenario depends on photochemical hazes forming and surviving in a hot, CO2-dominated atmosphere via roughly 1% H2S from volcanism, and the paper admits it is unclear whether such H2S abundances can be maintained, so if hazes cannot form the two-scenario conclusion reduces to a bare-surface interpretation.","fun_headline_variants_meta":{"raw":{"variants":["Two wavelengths, two faces: bare rock or hazy CO2","TRAPPIST-1 b's dual readings fit airless and hazy models","Bare rock or hazy CO2? Both fit TRAPPIST-1 b","New eclipses double TRAPPIST-1 b's possible worlds","Hazy CO2 atmosphere or bare rock? TRAPPIST-1 b keeps both"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00093,"raw_usage":{"total_tokens":4044,"prompt_tokens":1066,"completion_tokens":2978,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":682,"completion_tokens_details":{"reasoning_tokens":2876}},"tokens_in":682,"tokens_out":2978,"duration_ms":21215,"temperature":1.0,"reasoning_tokens":2876,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:45:21.304368+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A single 15 µm phase curve (the planned GO 3077 observations) would settle the matter: the airless ultramafic model predicts an almost vanishing nightside flux and a symmetric eclipse (night-to-day flux near zero), whereas the 10-bar hazy CO2 model predicts a night-to-day ratio of 0.85 and a phase-curve offset; if the measured nightside flux is near zero, the hazy-atmosphere scenario is ruled out.","supporting_citations":[],"review_version":1}