{"id":"04e8f558-a06f-4449-8168-28f77828e50a","arxiv_id":"2412.06648","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Three-dimensional cloud simulations show that crossing the Guillot humidity threshold in low-mean-molecular-weight atmospheres collapses the dry boundary layer into a superadiabatic cloudy layer.","lead":"This paper simulates thunderstorms and cloud layers in hydrogen-rich planetary atmospheres and finds that when the air near the surface is humid enough, the usual dry boundary layer collapses and is replaced by a very cloudy, extra-steep temperature layer. It suggests that young hydrogen-rich exoplanets may be systematically cloudier than expected, which could affect how we read their spectra.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The boundary-layer collapse is simulated under a fixed 200 W/m² uniform radiative cooling; if cloud–radiation feedback is added, the superadiabatic cloudy layer may not survive. No other load-bearing objection found.","rationale":"The reader's weakest assumption identifies the same issue: the prescribed 200 W/m² uniform radiative cooling is assumed not to qualitatively alter the convective regime transition. This is indeed the most load-bearing concern. The paper's own Sec. 3.1 and Discussion state that fixing the bulk tropospheric radiative cooling constrains the mean enthalpy flux and convective mass flux and that fully interactive radiative transfer is warranted. Because the proposed collapse mechanism in Sec. 4 depends on continued radiative cooling above the lifting condensation level, and because cloud fraction becomes nearly 100% at the transition, the positive feedback between cloudiness and reduced radiative cooling is a plausible route by which the collapse might not survive in a more realistic model. No internal inconsistency was found in the derivation of q_crit, and the high-resolution DeepBL_hr run provides some independent support for the simulated structure. The concern is significant enough that the paper should remain conditional rather than being accepted without qualification, but it is not a fatal flaw because the authors have clearly flagged the limitation and proposed the natural follow-up. Thus the reader's CONDITIONAL verdict is appropriate and should be kept unchanged.","tokens_in":13751,"tokens_out":6056,"duration_ms":73526,"concrete_test":"Re-run the VaryMu µa=2 case (or the DeepBL µa=4 case) with grey-gas interactive radiation, calibrating clear-sky cooling to the same 200 W/m² column integral in the non-Guillot reference state, and keep all other settings fixed. If the near-100% cloudy superadiabatic layer still emerges, the fixed cooling profile is not load-bearing; if the layer disappears or weakens substantially, the boundary-layer collapse is an artifact of the prescribed forcing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that crossing q_s ≥ q_crit (eq. 13) causes the dry subcloud layer to collapse into a cloudy superadiabatic layer. In the model, the column is forced by an idealized 200 W/m² cooling distributed uniformly between the surface and the diagnosed 200 K tropopause (Sec. 3.1). This fixes the total enthalpy flux and convective mass flux. The mechanism proposed in Sec. 4 explicitly relies on 'continued radiative cooling' above the lifting condensation level to amplify negative buoyancy. But the transition is also accompanied by cloud fraction jumping to nearly 100% (Fig. 5), so a fully interactive radiative scheme would likely cool much less in a cloudy, moist layer near the surface. If the cooling above the LCL is reduced by the clouds themselves, the runaway collapse described in Sec. 4 could be weakened or absent, and the 'very cloudy superadiabatic layer' could be an artifact of the prescribed forcing rather than a robust consequence of the Guillot threshold. The authors acknowledge this limitation and call for follow-up with fully interactive radiative transfer, but the claim as stated is conditional on it. The analytical derivation of q_crit itself appears sound and is not the source of the concern.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses the DAM cloud-resolving model to test a thermodynamic criterion first derived by Guillot (1995) for low-mean-molecular-weight atmospheres. The authors derive that when the saturation specific humidity exceeds qcrit = RT/[(μv-μa)L(T)], anomalously warm saturated air is negatively buoyant because the condensible's molecular weight outweighs the thermal expansion. In a suite of 3D radiative-convective equilibrium simulations with background molar masses from 28 to 2 g/mol, they find that crossing this threshold in near-surface humidity coincides with the disappearance of the dry subcloud layer and the formation of a shallow, nearly fully cloudy layer with temperature lapse rates well in excess of the dry adiabat. Additional suites with reduced surface wetness (DeepBL) show that a deep dry layer can keep the superadiabatic cloudy layer aloft, while increasing surface wetness collapses the boundary layer and sharply reduces column water vapor. The paper also reports episodic convection in one deep-boundary-layer case and computes 'cloud emission temperatures' for various condensibles at p = 0.1 bar.","tokens_in":14021,"tokens_out":9827,"duration_ms":104787,"significance":"If the main claim holds, the paper provides the first 3D cloud-resolving demonstration that a well-defined thermodynamic threshold controls convective structure in low-mean-molecular-weight atmospheres, with potentially observable implications for cloudiness and spectra of H2-rich planets. The analytical derivation is clean, the simulations cover a wide and thoughtfully designed parameter range, and the data and plotting code are made publicly available. The principal weakness is that the radiative forcing is idealized and prescribed, so the robustness of the boundary-layer collapse to cloud-radiation interactions remains untested; the authors acknowledge this and call for follow-up. With a focused sensitivity test or a qualified statement of conditionality, this would be a substantial contribution.","major_comments":[{"comment":"The mechanism for the boundary-layer collapse described in Section 4 explicitly relies on \"continued radiative cooling\" above the lifting condensation level to amplify the negative buoyancy of cloudy parcels. However, in all experiments the tropospheric radiative cooling is prescribed as a column-integrated 200 W/m2 distributed uniformly between the surface and the diagnosed 200 K tropopause (Section 3.1). Since the transition to the Guillot regime is accompanied by an increase in low cloud fraction to nearly 100% (Fig. 5, bottom), a fully interactive radiative transfer scheme would likely cool the cloudy layer less, which could weaken or eliminate the runaway collapse. The authors acknowledge this limitation in Sections 3.1 and 5, but the abstract and Section 4 state the collapse as a consequence of crossing the Guillot threshold without this caveat. I request a sensitivity test in which the vertical distribution of radiative cooling is varied (e.g., cooling reduced in the cloudy layer) or a demonstration that the qualitative regime transition is insensitive to the placement of the cooling; otherwise the headline claim is conditional on the prescribed forcing.","section":"Section 3.1 / Section 4"},{"comment":"The only higher-resolution run, DeepBL hr, uses a quasi-2D domain (750 m x 128 km) with a refined vertical grid, but it is presented only as a snapshot in Fig. 2 and is not compared quantitatively with its coarse-resolution counterpart (DeepBL with mu_a = 6 g/mol). The authors note that cloud fraction and precipitation efficiency are sensitive to horizontal and vertical resolution (Jeevanjee & Zhou 2022; Jenney et al. 2023), yet a central result is the jump of low cloud fraction to near 100% in the VaryMu runs (Fig. 5). A quantitative comparison of the time-mean cloud fraction, temperature profile, and layer depth between DeepBL hr and the coarse DeepBL mu_a = 6 run is needed to assess whether the near-100% cloudiness and the collapse of the subcloud layer are robust to resolution. If this comparison exists in the online material, it should be referenced explicitly in the main text.","section":"Section 3.1 / Table A1 / Fig. 2"}],"minor_comments":[{"comment":"In the inset of Fig. 3, the curves for different molar masses are plotted without an explicit line-style or color legend; please add a legend or a table of line styles so that the mu_a = 2, 4, and 6 cases can be distinguished.","section":"Fig. 3"},{"comment":"The two thin gray lines indicating qcrit for the 280 K and 320 K cases are almost indistinguishable from each other and from the dashed line for 300 K; consider using different colors and adding a legend.","section":"Fig. 5"},{"comment":"The sentence \"it is mixed down to the surface by turbulence and numerical diffusion\" is confusing: numerical diffusion should not be presented as a physical mixing mechanism. Please rephrase to describe turbulent mixing or model diffusion explicitly.","section":"Section 4"},{"comment":"For clarity, state explicitly that qcrit is a specific humidity (mass fraction) expressed in kg/kg, consistent with qv in Eq. (8); the table header already uses this unit, but the main text should do so as well.","section":"Section 2, Eq. (13)"}],"recommendation":"major_revision","confidential_remarks":"I found the paper strong in its qualitative conclusions and believe the main issue is the idealized forcing. I would suggest requiring the authors either to add a sensitivity test with a non-uniform or interactive radiation calculation, or to explicitly qualify the abstract by saying \"in simulations with prescribed radiative cooling.\" The high-resolution DeepBL run should be quantitatively compared to the coarse run; if that comparison is in a supplement, please make it explicit. The rest of the manuscript is well organized, and the data availability is a plus."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something useful: it takes Guillot's 1995 buoyancy-reversal threshold, re-derives it cleanly, and then shows in a systematic 3D cloud-resolving model that crossing q_s >= q_crit coincides with a sharp structural transition. The derivation (eqs. 5-13) is correct, and the simulations independently reproduce the transition without any fitted parameter. That is the main result, and it is solid.\n\nThe parameter sweep across mean molecular weight is the real novelty. Prior work was mostly 1D or focused on specific planets (K2-18b, Neptune); this paper maps the regime between Earth-like and H2-dominated convection. The DeepBL and VaryBeta experiments are a nice touch, showing that the superadiabatic cloudy layer can sit aloft with a deep dry subcloud layer or merge with the surface depending on moisture availability. The episodic convection in some runs is a bonus that will interest the cloud-dynamics community. Code and data are publicly available, which earns credit.\n\nWhere are the soft spots? The prescribed 200 W/m^2 radiative cooling, distributed uniformly between the surface and a diagnosed 200 K tropopause, is the big one. The stress-test concern is fair: when cloud fraction jumps to nearly 100%, a fully interactive radiation scheme would likely cool less in that cloudy layer, and the runaway collapse described in Section 4 could be weakened. The authors acknowledge this explicitly and call for follow-up with interactive radiation, so it is not a hidden flaw. It does mean the quantitative superadiabatic lapse rates and the near-100% cloud fraction should be treated as conditional on the idealized forcing. Also, there are no error bars on the time-mean profiles, and the resolution sensitivity is limited to one quasi-2D run. Those are minor-to-moderate issues.\n\nThe citation pattern is fair, and the paper situates itself well against Leconte et al. 2024 and Clément et al. 2024 without overselling novelty. The central argument holds up; the caveat is about how much of the extreme structure survives interactive radiation, not about whether the threshold matters.\n\nThis paper deserves serious peer review. A good referee will push for a sensitivity test with interactive radiation or at least a clearer discussion of how the fixed cooling profile may amplify the collapse. I would bring it to reading group and would cite it in my own work. Recommend accept with major revision.","headline":"A clean 3D confirmation that crossing Guillot's humidity threshold restructures the lower troposphere; the fixed radiative cooling is a real caveat but the paper flags it clearly and the core result stands.","tokens_in":14508,"tokens_out":1297,"would_cite":true,"duration_ms":15864,"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":"In hydrogen-rich atmospheres, crossing the Guillot humidity threshold makes the dry subcloud boundary layer collapse and be replaced by a near-100%-cloudy, superadiabatic layer.","keywords":["moist convection","hydrogen-rich atmospheres","Guillot threshold","buoyancy reversal","cloud-resolving model","superadiabatic lapse rate","sub-Neptune exoplanets","condensible species"],"falsifier":"A decisive test is to rerun the $\\mu_a = 2$ g/mol and $\\mu_a = 4$ g/mol cases with fully interactive radiative transfer: if the dry subcloud layer survives or the near-100% cloudy superadiabatic layer disappears, the collapse is an artifact of the prescribed cooling. A complementary observational check would be to look for the predicted cloud deck near the $\\sim 247$ K level in a solar-composition, H$_2$O-condensing sub-Neptune atmosphere.","tokens_in":13552,"feed_emoji":"☁️","tokens_out":9792,"duration_ms":92056,"temperature":0.7,"pith_summary":"On Earth, a warm, moist parcel of air is buoyant and rises. In hydrogen-rich atmospheres the condensible vapor (such as water) is heavier than the background gas, so a warm, saturated parcel can be denser than its surroundings and sink instead of rising. This paper uses a three-dimensional cloud-resolving model to show that this regime has a sharp onset: when near-surface specific humidity $q_s$ exceeds the critical value $q_{\\rm crit}$ first derived by Guillot (1995), the dry-adiabatic subcloud boundary layer collapses and is replaced by a shallow, extremely cloudy layer whose temperature drops faster than the dry adiabatic rate. With reduced surface moisture, the same superadiabatic cloudy layer forms aloft, above a deep dry layer, rather than at the surface. The result matters because hydrogen-rich atmospheres are common among sub-Neptunes and young planets, and the predicted increase in cloudiness may be observable.","feed_headline":"Hydrogen air makes warm clouds sink, simulations show","feed_subtitle":"Crossing the Guillot threshold collapses the dry boundary layer into a near-100% cloudy superadiabatic layer.","key_machinery":"The load-bearing object is the saturation density $\\rho^* = p \\mu^* / (RT)$ of a saturated air parcel, together with the critical humidity $q_{\\rm crit} = RT/[(\\mu_v - \\mu_a) L(T)]$ at which its logarithmic derivative with respect to temperature changes sign. When the vapor is heavier than the background gas ($\\mu_v > \\mu_a$), raising temperature raises the parcel's mean molar mass; above $q_{\\rm crit}$ that compositional effect beats thermal expansion, so anomalously warm saturated parcels are negatively buoyant. This identity supplies the mechanism for the simulated regime transition: condensation at cloud base produces parcels that sink rather than feed moist updrafts, so the dry subcloud layer erodes from the top until it either vanishes or is capped by a superadiabatic cloudy layer.","core_discovery":"The paper's central claim is that the Guillot humidity threshold is not just a thermodynamic curiosity but a dynamical regime boundary. In simulations with background molar mass $\\mu_a$ spanning 28 g/mol ($\\mathrm{N_2}$) down to 2 g/mol ($\\mathrm{H_2}$) at fixed surface pressure and temperature, the atmosphere's structure changes abruptly at the point where $q_s \\ge q_{\\rm crit}$ (eq. 13). Below the threshold, a dry subcloud layer overlies a moist troposphere close to the moist adiabat, with low cloud fractions of a few percent. Above it, the subcloud layer collapses: near-surface lapse rates become superadiabatic (about 120 K/km in the $\\mu_a = 2$ g/mol case), low cloud fraction jumps to nearly 100%, and the tropospheric temperature profile is much colder than a moist adiabat rooted at the surface. When surface moisture availability is reduced (the DeepBL suite), a deep dry subcloud layer survives and the cloudy superadiabatic layer forms aloft, with humidity falling by about an order of magnitude across it before the atmosphere returns to a second dry-adiabatic layer and then a moist layer. The paper interprets this as the expected consequence of negatively buoyant cloudy parcels at the lifting condensation level, and argues the trend toward increased cloudiness above threshold should apply to $\\mathrm{H_2O}$ and other condensible species.","pith_inferences":["We infer that the ratio $q_s/q_{\\rm crit}$ should organize cloud fraction across condensible species; running the same cloud-resolving model with CH$_4$ or NH$_3$ as the condensible would predict where superadiabatic decks form in gas-giant atmospheres.","We infer that with fully interactive radiation the transition may shift in parameter space, but the buoyancy-reversal mechanism should survive because it depends only on the thermodynamics of saturated parcels.","We infer that the episodic ~2-day convective pulses in some low-MMW runs suggest radiative-convective equilibrium in the Guillot regime may be bistable; global-scale simulations could reveal whether this manifests as observable, time-variable cloud cover.","We infer that if younger planets retain more H$_2$, the paper's cloudiness trend implies a secular brightening or dimming as hydrogen is lost and $\\mu_a$ rises; however, the paper notes the cloud decks in DeepBL sit below $p=0.1$ bar, so emission spectroscopy may be a more promising probe than transmission spectroscopy."],"forward_implications":["In atmospheres with $q_s \\ge q_{\\rm crit}$, the dry subcloud boundary layer collapses and low cloud fraction rises from a few percent to close to 100%, with the cloud peak dropping to the near-surface model level.","If surface moisture is limited ($\\beta=0.2$), the superadiabatic cloudy layer forms aloft above a roughly 30-km-deep dry subcloud layer, remaining distinct from both the surface layer and the moist troposphere; the paper calls this structure closer to sub-Neptune conditions.","Increasing surface moisture in the Guillot regime can paradoxically dry the atmosphere: in the VaryBeta suite, raising $\\beta$ from 0.4 to 0.6 collapses the boundary layer and cuts column water vapor from over 800 kg/m$^2$ to under 100 kg/m$^2$.","For solar-composition background gas ($\\mu_a=2.5$ g/mol) with water vapor, the superadiabatic cloudy layer would sit near $p=0.1$ bar at about 247 K, close to Earth's emission temperature, so such layers may influence the transmission and emission spectra of sub-Neptunes; the paper tabulates analogous cloud emission temperatures for other condensibles.","Some Guillot-regime simulations show episodic convection with roughly two-day pulses and associated variability in cloud water, indicating that convective activity in hydrogen-rich atmospheres can be intermittent."],"supporting_citations":[{"why":"Derives the critical humidity threshold $q_{\\rm crit}$ at which anomalously warm saturated air becomes negatively buoyant in a low-mean-molecular-weight background; the paper's central regime boundary.","marker":"Guillot (1995)"},{"why":"Provides the cloud-resolving model used for the three-dimensional simulations.","marker":"Romps (2008)"},{"why":"Supplies the experimental setup, surface flux formulae, and simplified microphysics scheme on which the present simulations build.","marker":"Seeley & Wordsworth (2023)"},{"why":"Introduces the surface wetness parameter $\\beta$ used in the DeepBL and VaryBeta suites to control moisture availability.","marker":"Cronin & Chavas (2019)"},{"why":"Provides the idealized moist-adiabat and scale-height comparison that the simulations are contrasted with, and motivates the low-MMW regime.","marker":"Koll & Cronin (2019)"},{"why":"Independent cloud-resolving simulations of K2-18b that produce qualitatively similar superadiabatic layers, used as a planetary-context comparison.","marker":"Leconte et al. (2024)"},{"why":"Theoretical study predicting superadiabatic layers aloft in sub-Neptunes, which the DeepBL experiments test and confirm.","marker":"Innes et al. (2023)"}],"fun_headline_variants":["Warm clouds sink in hydrogen atmospheres, simulations show","Humidity threshold in hydrogen air triggers near-total cloud cover","Guillot point: hydrogen atmospheres get superadiabatic cloud layers","3D runs show hydrogen atmospheres cloud up past Guillot humidity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes the prescribed 200 W/m$^2$ radiative cooling, spread uniformly between surface and tropopause, behaves enough like real radiative transfer that it does not qualitatively change the regime transition; if clouds or temperature structure feed back strongly on the cooling, the collapse and superadiabatic layer could be modified or absent.","fun_headline_variants_meta":{"raw":{"variants":["Warm clouds sink in hydrogen atmospheres, simulations show","Humidity threshold in hydrogen air triggers near-total cloud cover","Guillot point: hydrogen atmospheres get superadiabatic cloud layers","3D runs show hydrogen atmospheres cloud up past Guillot humidity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000302,"raw_usage":{"total_tokens":1815,"prompt_tokens":1093,"completion_tokens":722,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":709,"completion_tokens_details":{"reasoning_tokens":648}},"tokens_in":709,"tokens_out":722,"duration_ms":8580,"temperature":1.0,"reasoning_tokens":648,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:25:55.041908+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to rerun the $\\mu_a = 2$ g/mol and $\\mu_a = 4$ g/mol cases with fully interactive radiative transfer: if the dry subcloud layer survives or the near-100% cloudy superadiabatic layer disappears, the collapse is an artifact of the prescribed cooling. A complementary observational check would be to look for the predicted cloud deck near the $\\sim 247$ K level in a solar-composition, H$_2$O-condensing sub-Neptune atmosphere.","supporting_citations":[{"cited_title":"1995, Science, 269, 1697","cited_arxiv_id":null,"evidence_quote":"Derives the critical humidity threshold $q_{\\rm crit}$ at which anomalously warm saturated air becomes negatively buoyant in a low-mean-molecular-weight background; the paper's central regime boundary."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the cloud-resolving model used for the three-dimensional simulations."},{"cited_title":"W., & Chavas, D","cited_arxiv_id":null,"evidence_quote":"Introduces the surface wetness parameter $\\beta$ used in the DeepBL and VaryBeta suites to control moisture availability."},{"cited_title":"A 3D picture of moist-convection inhibition in hydrogen-rich atmospheres: Implications for K2-18 b","cited_arxiv_id":"2401.06608","evidence_quote":"Independent cloud-resolving simulations of K2-18b that produce qualitatively similar superadiabatic layers, used as a planetary-context comparison."}],"review_version":1}