{"id":"361d50b9-3de6-43f9-8ed6-5c9d310bf5d1","arxiv_id":"2501.04834","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Outflow emission from exoplanets is predicted to be most detectable in the Na I 589 nm doublet and the He* 1083 nm triplet, with HD 189733b the best candidate at a 410 ppm Na eclipse depth and SNR 2.4 per eclipse.","lead":"A new open-source 1D radiation-hydrodynamics code, pyTPCI, simulates escaping exoplanet atmospheres and predicts that the brightest outflow emission lines are the 589 nm sodium doublet and the 1083 nm metastable helium triplet. The models identify HD 189733b as the best target, with a sodium eclipse depth of 410 ppm and a per-eclipse SNR of 2.4, marginally within reach of Keck after 3 to 10 eclipses under optimistic photon-limited assumptions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The headline Na I signal for HD 189733b rests on the single 10x solar run that the paper itself flags as most violating the hydrodynamic assumption; at the better-fitting 1x solar metallicity the signal is ~16x weaker, so the viability claim is not yet established.","rationale":"We read the paper as making two nested claims: (1) pyTPCI can self-consistently model outflows and predict emission spectra, and (2) outflow emission is detectable, with HD 189733b as the best target. The code contribution is solid and the He* absorption validation for several planets gives some confidence in the framework. However, the specific detectability claim (410 ppm Na I, SNR 2.4) is not supported robustly. The signal is sourced entirely by the 10x solar metallicity run. The paper itself flags this run as the one that 'most badly violates the hydrodynamic assumption' and reports a mass-loss rate two orders of magnitude below the diffusion-limited value. At 1x solar, the Na I SNR is 0.15. Thus the predicted detectability hinges on an extrapolation to a metallicity regime where the code is unstable and where the physical outflow assumption is questionable. The reader's weakest assumption was XUV spectral uncertainty; while that is a genuine factor (Appendix A shows factor-of-a-few effects on He*), it does not have the potential to change the conclusion by an order of magnitude in the same way the metallicity choice does. If the 3-5x solar metallicity indicated by JWST/NIRCam yields a Na I SNR near 1 or below, the discovery potential is lost. This is not a question of current consensus but of internal consistency: the paper's own validation and stability statements undermine the single run that produces the headline. We therefore support the reader's conditional verdict but would shift the emphasis of the required revisions: run intermediate metallicities for HD 189733b, quantify the diffusion-limited condition with ionization, and report the Na I prediction for the 1x model prominently.","tokens_in":22204,"tokens_out":10599,"duration_ms":96504,"concrete_test":"Run pyTPCI for HD 189733b at 3x and 5x solar metallicity (the JWST/NIRCam range), with all other settings unchanged, and compute the Na I eclipse depth/SNR, He* absorption EW, and the ratio of simulated to ionization-corrected diffusion-limited mass-loss rate. If the 3-5x solar Na I SNR is <=1, or if the He* EW moves away from the observed 11 mÅ, or if the mass-loss rate remains below the diffusion limit, the claim that HD 189733b is a viable first emission target fails. As a sanity check, also re-derive the diffusion limit including the ionized fractions from the model; if the 10x run still violates it, the emission prediction is outside the code's domain.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central detectability claim—HD 189733b as a viable first target through 410 ppm Na I emission at SNR 2.4—is carried entirely by the 10x solar metallicity pyTPCI run (Table 2). The 1x solar run gives only 25 ppm depth and SNR 0.15, a factor ~16 lower. The 10x run predicts a He* absorption equivalent width of 7.5 mÅ versus the observed 11 mÅ, less close than the 1x value of 13 mÅ, so the available outflow diagnostic does not select the run that produces the headline signal. More seriously, Section 5.3 reports that the 10x solar HD 189733b simulation has a mass-loss rate of 1.9e8 g/s, about 200x below the diffusion-limited rate (3.8e10 g/s), and states that this run 'most badly violates the hydrodynamic assumption.' If the true outflow is not in the hydrodynamic regime, the CLOUDY emission calculation for that run is outside its domain of validity. The paper's speculation that ionization lowers the diffusion limit is not modeled. Meanwhile the JWST/NIRCam metallicity estimate of 3-5x solar (Fu et al. 2024) lies between the two simulated values, and no run at that metallicity is presented. The 10x result is therefore an extrapolation from a regime where the code is demonstrably unstable and inconsistent with the diffusion-limited condition. A plausible 1x solar outflow would make HD 189733b undetectable. This is the most load-bearing weakness because it collapses the specific discovery claim, independent of XUV uncertainties.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper introduces pyTPCI, an open-source Python reimplementation of TPCI that couples PLUTO and CLOUDY to simulate 1D radiatively driven atmospheric escape. The authors model seven planets at 0–100× solar metallicity, compute line emission spectra and eclipse depths/SNRs for the Na I 589 nm doublet, the He* 1083 nm triplet, Hα, and Mg I 457 nm for a 10 m-class high-resolution spectrograph, and validate the simulations against observed metastable helium absorption equivalent widths. Their headline result is that outflow emission is potentially detectable, with HD 189733b giving the strongest predicted signal: Na I eclipse depth 410 ppm and SNR 2.4 per eclipse in the 10× solar metallicity run, and He* depth 170 ppm and SNR 1.3 at 1× solar. They argue that such signals are marginally accessible with Keck in 3–10 stacked eclipses and considerably easier with next-generation ELTs.","tokens_in":22515,"tokens_out":9080,"duration_ms":85639,"significance":"The paper has clear strengths: pyTPCI is publicly released, Appendix A provides a direct code comparison with TPCI, and the simulations are benchmarked against an external observable—observed He* absorption equivalent widths—rather than being fitted to the emission predictions. If the predictions are correct, the paper opens a new observational window on escaping atmospheres and gives concrete target lists and line choices. The central caveat is that the quantitative discovery claim for HD 189733b rests on a single high-metallicity run that the authors themselves flag as violating the hydrodynamic assumption; the 1× solar run makes the same planet undetectable in Na I. The claim therefore needs either additional simulations in the observationally indicated metallicity range or a more conservative framing of what the existing runs can support.","major_comments":[{"comment":"The abstract's 'viable first target' statement is carried entirely by the HD 189733b 10× solar run. That run has a simulated mass-loss rate of 1.9×10^8 g/s, roughly 200× below the diffusion-limited rate of 3.8×10^10 g/s, and §5.3 states that it 'most badly violates the hydrodynamic assumption.' It is also not the run selected by the He* absorption benchmark: it predicts 7.5 mÅ versus the observed 11 mÅ, while the 1× solar run predicts 13 mÅ. The 1× run yields a Na I depth of only 25 ppm and SNR 0.15. Because Fu et al. (2024) infer 3–5× solar metallicity for this planet and no run at that metallicity is presented, the 410 ppm / SNR 2.4 value is an extrapolation across an unsimulated and unstable regime. Please either add the missing runs or explicitly demote the Na I detection claim.","section":"§5.3 and Table 2"},{"comment":"The discussion of the diffusion-limited mass-loss rate acknowledges that the neutral assumption may be wrong and that HD 189733b becomes highly ionized at low radii, but this argument is not quantified. The 10× solar run lies two orders of magnitude below the neutral diffusion limit, and the paper does not demonstrate that the reduction from ion–ion drag is large enough to make this run consistent with a hydrodynamic outflow. A quantitative estimate—for example, an effective diffusion limit computed with the simulated ion fractions, or a direct evaluation of the drag term—is needed before the 10× run can support a discovery claim.","section":"§5.3"},{"comment":"The predicted SNRs are not robust to the adopted stellar XUV spectra. Appendix A shows that switching from Salz et al. (2016) to Sanz-Forcada et al. (2025) spectra changes the HD 189733b 0× solar He* SNR from 1.7 to 0.76 and changes the WASP-107b 0× solar He* depth by roughly 50%; the text itself states that XUV fluxes are uncertain by factors of at least a few and that no EUV telescope is currently operating. The paper should present a systematic error budget on the headline 2.4 SNR and 410 ppm depth, or explicitly label those numbers as conditional on both the metallicity and the adopted XUV spectrum.","section":"Appendix A and §5.3"}],"minor_comments":[{"comment":"Please state explicitly that 'He Abs W' is the transit absorption percentage and that 'W' and 'W_obs' are equivalent widths in mÅ; the current column header is dense and easy to misread.","section":"Table 2 caption"},{"comment":"The introduction says that 'no work has yet considered whether outflows might also be detectable in emission,' but later cites Zhang et al. (2020) and refers to 'few published works' on the subject; please make the novelty claim consistent with the cited literature.","section":"Section 1"},{"comment":"The inset labeled 'Integrated Flux' reports values in ergs cm^-2 s^-1, while the y-axis is flux density per Å; please clarify the distinction or relabel the inset.","section":"Figure 2"},{"comment":"The notation δ(λ) is used for the maximum eclipse depth at the line peak, while the text also discusses a 'narrowband eclipse depth within ±2w'; please distinguish these two quantities explicitly in the equation and surrounding text.","section":"Section 4, Eq. (3)"}],"recommendation":"major_revision","confidential_remarks":"I would like the authors to respond to the first major comment before publication. The paper is otherwise well suited to the journal, and the code release plus external He* absorption validation are valuable. The issue is addressable by adding simulations at 3–5× solar metallicity for HD 189733b or by removing the 10× result from the headline claim, so I am not recommending rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth a look as a code release and as the first systematic emission-detectability study for exoplanet outflows, but the headline claim about HD 189733b is not yet supported.\n\nWhat's new: pyTPCI is an open-source rewrite of TPCI with modern PLUTO/CLOUDY, easier installation, and better stability. The Appendix A benchmark against Salz's TPCI for HD 189733b and HD 209458b is genuinely reassuring. The paper then forward-models emission spectra for seven planets at multiple metallicities, computes eclipse depths and SNRs, and checks absorption equivalent widths against observations. For four systems the He* widths are reproduced well. That is real, reproducible work, and the code release is a service to the community.\n\nSoft spots: the Na I SNR of 2.4 and 410 ppm depth come from the 10x solar run that the paper itself says \"most badly violates the hydrodynamic assumption\" (Section 5.3), with a mass-loss rate about 200x below the diffusion-limited rate. At 1x solar the same planet gives SNR 0.15 and 25 ppm. The observed He* absorption lies between the 1x and 10x predictions, and JWST/NIRCam suggests 3-5x solar, which is not simulated. So the specific discovery claim rests on an extrapolation out of the code's valid regime. Three of the seven systems are also discarded because the model overpredicts He* absorption, which narrows the validation set. XUV flux uncertainty is acknowledged, and Appendix A shows SNRs changing by up to 50% when the input spectra are swapped. These are real limitations, and they land on the central claim.\n\nCredit where earned: the paper is transparent about its limitations, the emission predictions are genuinely forward-modeled (no circular fitting to emission data), and the line-width, optical-depth, and photosphere-radius calculations are clearly laid out. The authors frame this as an order-of-magnitude investigation, which is the right frame.\n\nBottom line: the code and the question deserve referee time, but the flagship detectability claim needs support it does not yet have. A run at the JWST-inferred metallicity, error propagation on depths and SNRs, or a physical argument for why the 10x run survives the diffusion-limit concern would materially improve the paper. As it stands, this is a conditional accept at most, and I would push for major revision.","headline":"A valuable code release and a first systematic look at outflow emission detectability, but the flagship HD 189733b Na I claim rests on a metallicity run the authors themselves flag as violating the hydrodynamic assumption.","tokens_in":23070,"tokens_out":1870,"would_cite":false,"duration_ms":20248,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Photoevaporating exoplanet atmospheres can be detected in emission, not only in absorption during transit.","keywords":["exoplanet atmospheres","photoevaporation","atmospheric escape","outflow emission","metastable helium","sodium doublet","radiative-hydrodynamics","pyTPCI"],"falsifier":"Take high-resolution spectra of HD 189733b during several secondary eclipses with a 10-meter telescope and search for the predicted 589 nm Na I line at 410 ppm depth with SNR 2.4 per eclipse; if five to ten stacked eclipses show no line at the predicted depth, the outflow emission model or the assumed XUV ionization is wrong. A cleaner test would be to obtain direct stellar EUV spectra and re-run the pyTPCI models to see whether the predicted sodium depth survives.","tokens_in":21961,"feed_emoji":"🪐","tokens_out":8180,"duration_ms":70722,"temperature":0.7,"pith_summary":"This paper asks whether gas escaping from irradiated exoplanets can be observed in emission rather than only in absorption during transit. To answer it, the authors present pyTPCI, a new open-source one-dimensional radiation-hydrodynamics code that couples the PLUTO hydrodynamics solver to the CLOUDY plasma microphysics code, and use it to model seven planets at several metallicities. They compute eclipse depths and signal-to-noise ratios for the outflow's spectral lines as seen by a 10-meter telescope with a high-resolution spectrograph. The central prediction is that the 589 nm Na I doublet and the 1083 nm metastable helium triplet are the most detectable lines, and that HD 189733b is the most promising target, with a Na I eclipse depth of 410 ppm and SNR of 2.4 per eclipse, marginally detectable with a 10-meter telescope after stacking 3 to 10 eclipses under photon-limited assumptions. The intended point is that outflows open a new emission-based window on escape processes and the radius valley they sculpt.","feed_headline":"Exoplanet outflows may be detectable in emission, not just absorption","feed_subtitle":"Seven-planet grid points to HD 189733b's sodium and helium outflow lines, reachable by stacking a few eclipses.","key_machinery":"The load-bearing object is pyTPCI, a 1D radiation-hydrodynamics code coupling PLUTO's hydrodynamic solver to CLOUDY's photoionization and spectral synthesis. CLOUDY computes the heating, cooling, ionization, and line emission from a specified stellar XUV spectrum; PLUTO evolves density, pressure, and velocity; and the two are iterated to a converged outflow. The observability calculation then uses the CLOUDY spectrum: each line is broadened by thermal, wind, rotational, and instrumental widths, the radius where the line's optical depth reaches unity sets the effective emitting area (the photosphere radius), and the eclipse depth is the ratio of planet to stellar surface flux times the squared photosphere-to-star radius ratio. The code also reproduces the observed metastable helium absorption equivalent widths for four of the seven planets, which the authors use as a validation check on the underlying outflow structure.","core_discovery":"The paper's claim is that exoplanet outflows are not just absorption features; under realistic 1D radiation-hydrodynamic modeling they emit enough flux in narrow lines to be observable at eclipse. For the seven simulated planets, the strongest and most consistently detectable emission lines are the Na I doublet near 589 nm and the metastable helium triplet at 1083 nm, with Halpha and the 457 nm Mg I line appearing occasionally but usually optically thin and therefore not clearly attributable to the outflow. Among all models, HD 189733b at 10x solar metallicity gives the highest signal: a Na I eclipse depth of 410 ppm and SNR of 2.4 per eclipse, plus a He* eclipse depth of 170 ppm and SNR of 1.3 from the 1x solar run. The authors argue that with 3 to 10 eclipses these signals would be marginally detectable with a 10-meter telescope, and that a next-generation 40-meter telescope would push the Na signal to a SNR around 10.","pith_inferences":["If the predicted Na I emission is real, a failed search at the quoted depth would itself be informative: it would point to an XUV flux or ionization balance different from the assumed one, since neutral sodium density is the controlling factor.","The same detectability machinery could be applied to planets around M dwarfs, which the paper identifies as untried but potentially favorable because of their large planet-to-star radius ratios.","A time-resolved emission measurement during eclipse would let observers watch the outflow accelerate as it escapes; absorption measurements only give the integrated column, so emission adds kinematic structure.","The paper's line-broadening formula predicts specific line widths, so comparing observed line shapes to those predictions would directly test the assumed outflow temperature and velocity profiles."],"forward_implications":["The 589 nm Na I doublet and the 1083 nm metastable helium triplet are the best observational targets for outflow emission searches, ahead of Halpha and Mg I.","HD 189733b is the most promising first target: three to ten stacked eclipse observations with a 10-meter telescope could reach its predicted Na I signal if the noise is photon-limited.","Emission detections would supply independent constraints on outflow temperature, density, ionization state, and metallicity, complementing transit absorption measurements.","A 40-meter class telescope would push the predicted Na I SNR for HD 189733b to roughly 10, making robust emission studies feasible.","Higher metallicity strengthens Na I emission up to a point, but at 100x solar cooling suppresses the outflow and weakens the line."],"supporting_citations":[{"why":"Introduced TPCI, the PLUTO-CLOUDY interface that pyTPCI improves, supplying the base coupling method.","marker":"Salz et al. (2015)"},{"why":"Described TPCI simulations of escaping atmospheres and the stellar spectrum construction method used for comparison runs, providing benchmarks.","marker":"Salz et al. (2016)"},{"why":"Established metastable helium as the exosphere tracer whose absorption measurements are used for validation.","marker":"Oklopčić & Hirata (2018)"},{"why":"Provided the EMD stellar XUV spectra (1-912 Angstrom) that set the ionization balance and drive the outflows in the production runs.","marker":"Sanz-Forcada et al. (2025)"},{"why":"Detected helium absorption from WASP-107b, motivating that simulated target and providing an observed equivalent width.","marker":"Spake et al. (2018)"},{"why":"Detected helium absorption from WASP-69b, contributing a target and observed equivalent width.","marker":"Nortmann et al. (2018)"},{"why":"Observed helium absorption for TOI-560b and TOI-1430b, the mini-Neptune targets whose 100x solar models match the observed widths.","marker":"Zhang et al. (2023)"},{"why":"Transmission spectra indicating HD 189733b has a metallicity of 3-5x solar, supporting the 1x and 10x solar models that best match its helium absorption.","marker":"Fu et al. (2024)"}],"fun_headline_variants":["Exoplanet outflows emit detectable sodium and helium lines","HD 189733b outflow shines in sodium and helium emission","Outflow emission detectable? Try sodium and helium lines","Sodium and helium emission from exoplanet outflows now forecast","Exoplanet outflows might glow in sodium and helium"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The predictions assume the assumed stellar XUV spectra correctly set the outflow's ionization balance and metal line populations; those fluxes are uncertain by factors of at least a few, no EUV telescope is currently operating, and the sodium emission in particular depends on the density of neutral sodium set by that ionization state.","fun_headline_variants_meta":{"raw":{"variants":["Exoplanet outflows emit detectable sodium and helium lines","HD 189733b outflow shines in sodium and helium emission","Outflow emission detectable? Try sodium and helium lines","Sodium and helium emission from exoplanet outflows now forecast","Exoplanet outflows might glow in sodium and helium"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000222,"raw_usage":{"total_tokens":1547,"prompt_tokens":1131,"completion_tokens":416,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":747,"completion_tokens_details":{"reasoning_tokens":333}},"tokens_in":747,"tokens_out":416,"duration_ms":4128,"temperature":1.0,"reasoning_tokens":333,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:24:02.702000+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take high-resolution spectra of HD 189733b during several secondary eclipses with a 10-meter telescope and search for the predicted 589 nm Na I line at 410 ppm depth with SNR 2.4 per eclipse; if five to ten stacked eclipses show no line at the predicted depth, the outflow emission model or the assumed XUV ionization is wrong. A cleaner test would be to obtain direct stellar EUV spectra and re-run the pyTPCI models to see whether the predicted sodium depth survives.","supporting_citations":[],"review_version":1}