{"id":"7350a8d1-6f20-419c-b019-add822da459f","arxiv_id":"1908.03510","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Close-in giant planets lose mass fastest when young: a 0.3-Jupiter-mass planet can lose up to 20% of its mass, a 1-Jupiter-mass planet loses under 1%, and hydrogen-alpha transit signals fade after roughly 1.2 billion years.","lead":"This paper simulates how close-in giant planets lose their atmospheres over billions of years and predicts what hydrogen lines should look like during transits. It finds that a lower-mass giant can lose up to a fifth of its mass, which may explain the Neptunian desert, and that hydrogen-alpha signals fade after about a billion years.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"20% mass-loss claim and 1.2-Gyr H-alpha boundary rest on quoted radius tracks, but the sign and magnitude of the neglected mass-feedback are not established; the paper's own stated limitation should be quantified before the Neptunian-desert conclusion is used.","rationale":"The reader's weakest_assumption is exactly the one I find load-bearing: the fixed-mass, no-mass-loss radius tracks used in the integration. The paper is transparent about this simplification, but transparency does not make the assumption safe for the central quantitative claim. The 20% mass-loss fraction for the 0.3-Mjup planet is large enough that ignoring the mass decrease is not a priori negligible; the paper's statement that the effect is minor is an assertion, not a result. The H-alpha disappearance at ~1.2 Gyr depends on the n=2 population in the extended atmosphere, which depends on temperature and density structure of the wind, which in turn depends on the assumed mass and radius. Thus the same fix affects both headline numbers. The reader identifies the same weakness and recommends a conditional verdict; I concur. I do not see an internal inconsistency in the derivations as written—the equations and the numerical setup are self-consistent given the stated inputs. The concern is about the applicability of the model to the evolutionary claim, which is a correctness risk rather than a methodological fraud. A concrete self-consistent re-run or even a sensitivity test (e.g., reducing Mpl by 10–20% at late ages and checking the change in Mdot and H-alpha depth) would settle whether the 20% figure and the 1.2-Gyr boundary shift materially. This is a standard check and is feasible with the described code. The verdict should remain CONDITIONAL, with the condition being that the mass-loss feedback be quantified.","tokens_in":18795,"tokens_out":1813,"duration_ms":16845,"concrete_test":"Re-run the evolutionary integration for the 0.3-Mjup fast-rotator track with mass updated self-consistently: at each age step, compute Mdot with the current mass (and with radius re-scaled approximately for the lower mass, e.g., using the Fortney & Nettelmann radius-mass relation), reduce Mpl accordingly, and recompute the H-alpha population and transit. If the total mass lost stays below 20% and H-alpha still vanishes near 1.2 Gyr, the concern is minor. If mass lost exceeds 20% significantly or the H-alpha age boundary shifts by more than a factor of two, the paper's headline claims need revision and the Neptunian-desert conclusion becomes conditional on self-consistent evolution.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claims—0.3-Mjup losing up to 20% of its mass and H-alpha disappearing after ~1.2 Gyr—are derived by integrating a mass-loss rate Mdot(t) that the paper itself states is computed with a fixed planetary mass and with radius evolution from Fortney & Nettelmann (2010) 'that do not account for the mass loss' (Section 2), and 'We do not include this decrease in mass in our simulations' (Section 4). The paper's defense, that the mass variation effect is minor compared to radius variation (Section 4), is not demonstrated. Mdot in this model scales roughly as a power of the effective gravitational potential GMpl/Rpl; if the planet loses ~20% of its mass while its radius evolution is that of a 0.3-Mjup object, its gravitational potential is overestimated at late ages. The direction of the effect is clear: escape rates would be larger, fractions larger, and the 1.2-Gyr H-alpha boundary could shift to older ages. Quantitatively, the claimed 20% could be a lower limit, and the conclusion that the Neptunian desert is attributable to mass loss would be strengthened or weakened depending on how self-consistent evolution changes Mdot. The paper's own text flags this as a limitation, but the headline claim is used to support a physical conclusion about the Neptunian desert. Without a sensitivity or self-consistent calculation, the 20% figure and the age boundary are not robust.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents one-dimensional hydrodynamic escape simulations for 0.3- and 1-Mjup planets at 0.045 au around a solar-mass star, evolved from 10 Myr to 5 Gyr under three stellar EUV evolution tracks (slow, intermediate, fast rotators). The authors compute mass-loss rates, integrate them to obtain total mass loss, and then use a three-dimensional ray-tracing model with Voigt profiles to predict Ly-alpha and H-alpha transit depths and equivalent widths. The main claims are that the 0.3-Mjup planet can lose up to 20% of its initial mass, supporting a mass-loss origin of the Neptunian desert, and that H-alpha absorption nearly disappears after about 1.2 Gyr because the outflowing hydrogen is predominantly in the ground state.","tokens_in":19002,"tokens_out":6470,"duration_ms":65532,"significance":"If the quantitative results hold, the paper provides an interesting evolutionary framework: it shows that escape can be dynamically important for low-gravity giants, gives specific predictions for H-alpha and Ly-alpha transit evolution, and offers analytical fits to mass-loss rates as an alternative to energy-limited formulas. The model is transparent in its limitations (4-pi isotropic upper limit, fixed mass), and the comparison with the energy-limited approximation is instructive. The predictions are falsifiable with observations of young systems, and the inclusion of three stellar rotation tracks is a strength. However, because the headline mass-loss fraction and the H-alpha disappearance age rest on the unquantified fixed-mass assumption, the robustness of these conclusions is not yet established.","major_comments":[{"comment":"The integrated total mass loss of more than 20% for the 0.3-Mjup planet is obtained by integrating mass-loss rates that are computed using a fixed planetary mass and radius tracks from Fortney & Nettelmann (2010) that do not include mass loss, as stated in Sections 2 and 4. For a planet that loses more than 20% of its mass, the gravitational potential at late ages is overestimated, and the paper's assertion that this effect is 'minor' compared to radius variation (Section 4) is not demonstrated. Since the mass-loss rate in this model scales with the effective gravitational potential, the self-consistent mass-loss fraction could be substantially larger. The authors should quantify this by a sensitivity test, for example by repeating the evolution with a mass that decreases according to the computed Mdot, or by providing an analytic estimate of the fractional change in Mdot when Mpl is reduced by 20%.","section":"Section 4 and 4.1"},{"comment":"The prediction that H-alpha absorption nearly disappears after ~1.2 Gyr for the 0.3-Mjup planet is derived from atmospheric structures computed with the same fixed-mass approximation. Because the planet has already lost a significant fraction of its mass by that age in the authors' own integration, the late-age wind density, temperature, and n=2 population may be misrepresented. The 1.2-Gyr boundary should therefore be presented as a lower limit or as model-dependent until the mass-feedback effect is assessed; as written, this prediction is not robust to the stated limitation.","section":"Section 5.2, Figure 8"}],"minor_comments":[{"comment":"The Voigt profile expression includes a variable xi that is never defined; please define it (or remove it) for completeness.","section":"Section 3.3, Eq. (13)"},{"comment":"The normalization of the Johnstone et al. (2015c) EUV curves by dividing by 6.7 is described in a footnote; it would be clearer to state in the main text that the entire evolutionary track is scaled by a constant factor to match the Ribas et al. (2005) solar-age value.","section":"Section 2"},{"comment":"The comparison to HD189733b and KELT-9b would benefit from a brief statement of the model ages or EUV fluxes that correspond to the quoted observed transit depths, since the paper's own predictions are age-dependent.","section":"Section 5.2"},{"comment":"The abstract says H-alpha absorption 'nearly disappears' after ~1.2 Gyr; Figure 8 shows that for the 0.3-Mjup planet the excess flux goes to zero, while for the 1-Mjup planet it remains at fractions of a percent. The wording should distinguish these cases.","section":"Abstract and Figure 8"}],"recommendation":"major_revision","confidential_remarks":"The authors explicitly acknowledge the fixed-mass limitation, which is good scientific practice. However, the Editors may wish to require a quantitative estimate of the effect before publication, given that the 20% figure and the H-alpha disappearance age are headline results. I have no concerns about the novelty or scope; the paper fits the journal well."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a workmanlike extension of the Murray-Clay hydrodynamic escape model to an evolutionary grid: two planet masses, three stellar rotation tracks, 10 Myr to 5 Gyr, with Ly-alpha and H-alpha transit predictions at every step. The main new deliverables are the time-dependent mass-loss tracks and the analytical fits in Appendix A, plus the prediction that H-alpha absorption fades as the planet ages. That qualitative picture is credible and useful for observers.\n\nWhat the paper does well: it runs a large number of models, compares against the energy-limited approximation in a way that shows where that approximation breaks, and is transparent about its assumptions. The comparison to HD189733b and GJ436b is sensible. The transit code upgrade to Voigt profiles is incremental but reasonable. The authors clearly state that the radius tracks do not include mass loss and that the wind is an isotropic 4-pi upper limit.\n\nThe soft spots are real, and the stress-test note lands. The 20% mass-loss figure for the 0.3-Mjup planet is computed by integrating Mdot while holding the planet mass fixed. Since Mdot in this model increases when the gravitational potential weakens, a self-consistent treatment would give a larger integrated loss, not a smaller one. So the paper's headline number is a lower limit, and the Neptunian-desert conclusion would likely be strengthened, not threatened, by including mass feedback. But the authors do not quantify this, and the statement that the effect is minor compared to radius variation is asserted, not demonstrated. A referee should ask for a sensitivity test or a simple self-consistent integration.\n\nThe H-alpha disappearance at ~1.2 Gyr has a second soft spot: the n=2 population is computed in the coronal-model approximation with only direct excitation from the ground state, no radiative pumping, and no collisional de-excitation. The paper notes that Boltzmann equilibrium gives up to a factor of two larger transits, which suggests the age boundary is not robust either. Again, this is acknowledged, but it is a caveat on the headline claim.\n\nNo code or data are shipped, so independent reproduction is possible but nontrivial. That said, the paper is honest about its limitations, and the central qualitative conclusions hold. I would send this to peer review and ask for the mass-feedback sensitivity test before accepting the quantitative values. The paper is worth a read for anyone modeling or observing escaping exoplanet atmospheres, and the analytical fits alone are a useful citable contribution.","headline":"Useful evolutionary escape predictions with a real caveat: the headline 20% and 1.2-Gyr boundary are lower limits, not robust values.","tokens_in":19663,"tokens_out":2450,"would_cite":true,"duration_ms":29743,"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":"A close-in Saturn-mass giant can lose 20% of its mass to escape, while a Jupiter twin loses at most 1%.","keywords":["atmospheric escape","hydrodynamic escape","Ly-alpha transits","H-alpha transits","Neptunian desert","hot Jupiters","planet evolution","EUV irradiation"],"falsifier":"Observe a 0.3-Jupiter-mass planet near 0.045 au around a solar-type star older than about 1.2 Gyr and measure its H-alpha transit: the model predicts essentially zero excess absorption, so a secure detection of more than 1% excess would contradict the ground-state domination prediction. Conversely, finding a substantial population of such planets at old ages would weaken the claim that escape removes most of their mass.","tokens_in":18443,"feed_emoji":"🪐","tokens_out":8540,"duration_ms":79629,"temperature":0.7,"pith_summary":"The paper simulates how atmospheric escape evolves for $0.3$ and $1$ Jupiter-mass planets orbiting a solar twin at $0.045$ au from ages $10$ Myr to $5$ Gyr, then turns the simulated outflows into Ly$\\alpha$ and H$\\alpha$ transit light curves. Its central result is that escape matters for low-gravity giants: the $0.3\\,\\mathrm{M_{Jup}}$ planet can lose more than $20\\%$ of its initial mass over its life, whereas the $1\\,\\mathrm{M_{Jup}}$ planet loses at most $1\\%$. That asymmetry is presented as direct support for the idea that the short-period Neptunian desert is carved by mass loss. The paper also predicts that H$\\alpha$ transits are strong only in young systems, reaching $3\\text{--}4\\%$ excess depth, and nearly vanish after about $1.2$ Gyr because the escaping neutral hydrogen sits almost entirely in the ground state. A sympathetic reader would take this as a roadmap for where H$\\alpha$ surveys should look.","feed_headline":"Escape can strip 20% off a close-in Saturn","feed_subtitle":"Model tracks giants from 10 Myr to 5 Gyr and predicts H-alpha transits vanish by 1.2 Gyr.","key_machinery":"The carrying mechanism is a one-dimensional, spherically symmetric hydrodynamic escape model for a hydrogen atmosphere, with photoionization heating from stellar EUV, Ly$\\alpha$ radiative cooling, ionisation balance, and conservation of mass, momentum (including a tidal term), and energy; the transonic outflow is solved by a shooting method. Two evolutionary inputs drive the time dependence: stellar EUV flux following slow, intermediate, and fast rotator tracks, and radius-contraction curves for the $1$ and $0.3$ Jupiter-mass planets. For transits, the outflow density, temperature, and velocity are mapped onto a 3D grid and ray-traced with Voigt line profiles for Ly$\\alpha$ and H$\\alpha$, with the $\\mathrm{n}=2$ hydrogen fraction computed from statistical equilibrium; this is what produces the grounded prediction that H$\\alpha$ absorption nearly disappears after about $1.2$ Gyr.","core_discovery":"On its own terms, the paper claims that the mass-loss history of a close-in giant is set by two competing evolutions: stellar EUV flux declines with age (faster for slow rotators), while the planet contracts and its gravity strengthens. Combining these inputs in a one-dimensional hydrodynamic wind model yields mass-loss rates of roughly $10^9$--$10^{12}$ g/s for the Jupiter-mass planet and $10^{10}$--$10^{13}$ g/s for the $0.3$ Jupiter-mass planet, declining steeply with age. Integrating these rates gives the headline numbers: at most $1\\%$ of the initial mass lost for the $1\\,\\mathrm{M_{Jup}}$ case and more than $20\\%$ for the $0.3\\,\\mathrm{M_{Jup}}$ case, with the exact fraction depending on whether the host star was born a slow or fast rotator. The same outflows, rendered with ray tracing, produce saturated Ly$\\alpha$ at line center for young planets and H$\\alpha$ excesses of at most $3\\text{--}4\\%$ that fade to nothing after about $1.2$ Gyr, because the $\\mathrm{n}=2$ hydrogen population needed for H$\\alpha$ becomes negligible while the ground-state population that drives Ly$\\alpha$ remains large.","pith_inferences":["If mass loss is as strong as claimed for the 0.3-Jupiter-mass case, then including the lost mass in the radius evolution would weaken gravity at late ages and likely push the total lost fraction above 20%; the quoted numbers should be read as lower bounds.","The same machinery, run for lower-mass planets or shorter orbits, could quantify whether some close-in rocky planets are fully stripped cores of former Neptunes, an endpoint the paper explicitly gestures toward.","Because stellar activity mimics H-alpha variability, the model's prediction of stronger H-alpha at young ages could be tested by targeting young, inactive stars in clusters precisely where activity is lowest.","The predicted disappearance of H-alpha after about 1.2 Gyr offers a direct observational test of EUV evolution models: a detected H-alpha excess around an old, low-gravity giant would require either higher EUV than assumed or additional excitation mechanisms."],"forward_implications":["Over 5 Gyr, a Jupiter-mass planet at 0.045 au keeps at least 99% of its mass, so hot-Jupiter mass loss is negligible for the planet's overall evolution.","A 0.3-Jupiter-mass planet at the same orbit can lose more than a fifth of its mass, making hydrodynamic escape a viable sculptor of the Neptunian desert.","H-alpha transits are a young-planet phenomenon: excess depths of 3-4% occur only at the youngest ages, and excess absorption drops below 1% after roughly 200 Myr (or 600 Myr around fast rotators).","After about 1.2 Gyr, H-alpha absorption vanishes even where Ly-alpha remains strong or saturated, so joint H-alpha and Ly-alpha observations can distinguish young escaping atmospheres from old quiescent ones.","The power-law fits for mass-loss rate and terminal velocity let observers estimate escape rates from age and host-star rotation alone, without recomputing hydrodynamics."],"supporting_citations":[{"why":"Provides the hydrodynamic escape model (heating, cooling, ionisation balance, transonic wind solution) on which the evolution runs are based.","marker":"Murray-Clay et al. (2009)"},{"why":"Supplies the radius-contraction curves for the 1- and 0.3-Jupiter-mass planets, which set the planetary gravity at each age.","marker":"Fortney & Nettelmann (2010)"},{"why":"Supplies the evolving EUV flux tracks for slow, intermediate, and fast rotators that drive the irradiation level over time.","marker":"Johnstone et al. (2015c)"},{"why":"Provides the ray-tracing transit model, adapted here with Voigt profiles, that converts the outflow into Ly-alpha and H-alpha transits.","marker":"Vidotto et al. (2018)"},{"why":"Defines the energy-limited escape formula against which the hydrodynamic mass-loss rates are compared.","marker":"Watson et al. (1981)"},{"why":"Gives the tidal correction factor used in the corrected energy-limited mass-loss estimates.","marker":"Erkaev et al. (2007)"},{"why":"Documents the short-period Neptunian desert that the paper's 20% mass-loss result is invoked to explain.","marker":"Mazeh et al. (2016)"},{"why":"Provides the atomic data and level-population routines used to compute the n=2 hydrogen density for H-alpha transits.","marker":"Dere et al. (2019)"}],"fun_headline_variants":["Close-in giants lose up to 20% mass, H-alpha fades by 1.2 Gyr","Escape evolution strips 20% off low-mass giants","Young giants escape fast, H-alpha transits vanish with age","How close-in giants lose mass and their H-alpha signal","Mass loss models tie Neptunian desert to giant evolution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The radius and mass evolution are taken from contraction tracks that do not include the mass being lost, so for the 0.3-Jupiter-mass planet the gravity used at late ages is too strong; including the lost mass would raise escape and could change the quoted 20% and 1.2 Gyr H-alpha boundary.","fun_headline_variants_meta":{"raw":{"variants":["Close-in giants lose up to 20% mass, H-alpha fades by 1.2 Gyr","Escape evolution strips 20% off low-mass giants","Young giants escape fast, H-alpha transits vanish with age","How close-in giants lose mass and their H-alpha signal","Mass loss models tie Neptunian desert to giant evolution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000596,"raw_usage":{"total_tokens":2894,"prompt_tokens":1158,"completion_tokens":1736,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":774,"completion_tokens_details":{"reasoning_tokens":1642}},"tokens_in":774,"tokens_out":1736,"duration_ms":13303,"temperature":1.0,"reasoning_tokens":1642,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:11:32.422256+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Observe a 0.3-Jupiter-mass planet near 0.045 au around a solar-type star older than about 1.2 Gyr and measure its H-alpha transit: the model predicts essentially zero excess absorption, so a secure detection of more than 1% excess would contradict the ground-state domination prediction. Conversely, finding a substantial population of such planets at old ages would weaken the claim that escape removes most of their mass.","supporting_citations":[{"cited_title":"A., et al., 2018, @doi [ ] 10.1093/mnras/sty2130 , http://adsabs.harvard.edu/abs/2018MNRAS.481.5296V 481, 5296","cited_arxiv_id":null,"evidence_quote":"Provides the ray-tracing transit model, adapted here with Voigt profiles, that converts the outflow into Ly-alpha and H-alpha transits."}],"review_version":1}