{"id":"2a8cd93a-a936-4b28-8075-590ff9290b91","arxiv_id":"2506.23891","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"On K2-18b, a 3D climate model shows atmospheric winds concentrate long-lived gases at the evening terminator, and the fastest rotation studied creates tracer-rich polar regions.","lead":"Researchers ran 3D climate simulations of the temperate sub-Neptune K2-18b with a passive tracer to see how winds move gases. They find evening-side enrichment of long-lived gases and strong rotation-dependent transport at high latitudes, which matters for reading JWST spectra.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Conserved passive tracer with no source/sink has no steady state; the reported 20% terminator asymmetry and Kzz may be transient artifacts of the Eq. (4) initial profile and run length.","rationale":"The reader's weakest-assumption identification is on target: Eq. (4) is load-bearing and untested. I agree with the conditional verdict. The paper has real strengths — the zonal momentum budget (Eq. 7, Figs. 6–7) is a careful diagnostic, conservation is checked (Appendix A), and the two-paper structure is honest about deferring active chemistry. My concern is narrower: the conserved tracer has no steady state, so every tracer-derived number has an implicit dependence on initialization and run length. The active-species comparison uses data already in hand to settle whether this dependence actually changes the conclusions. I do not see an internal inconsistency in the dynamics; the issue is an unsupported premise for the transport claim. I also considered the composition-tuning-to-JWST issue, but that is an external degeneracy rather than an internal error, and the paper states its assumptions clearly. Thus no verdict change: CONDITIONAL remains appropriate, with the sensitivity test as the natural condition.","tokens_in":30767,"tokens_out":9307,"duration_ms":115723,"concrete_test":"Re-analyze the already-computed kinetics runs before any new simulations: calculate the evening-minus-morning terminator contrast and the Eq. (8) Kzz profile for the modeled CH4 and CO2 fields (the planned Part II analysis) and compare them with the passive-tracer results. If CH4 and CO2 reproduce the ~20% asymmetry and the same Kzz shape, the no-source passive tracer is a valid proxy and the concern does not land. If the active species show a different limb contrast or Kzz, the conserved-tracer setup is controlling the headline numbers.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption is the passive-tracer design in §2.3. Equation (4) initializes a conserved tracer (dq/dt=0) with q=1e-5 below 10 bar and a steep (P/10 bar)^1.5 decrease above, and no source, sink, or upper-boundary loss is applied. For a closed domain with no sources or sinks, the only steady solution of the advection-diffusion equation is a globally uniform mixing ratio; the 'quasi-steady state' invoked to justify the analysis is therefore a transient set by the initial gradient and the vertical mixing timescale. Figure B2(f) shows the tracer at 0.001 bar evolving over the 5100-day kinetics run, but no plateau criterion is specified or quantified. Consequently, the ~20% evening/morning terminator contrast (Fig. 11) and the equivalent 1D Kzz profile (Fig. 8, Eq. 8) measure the erosion of the arbitrarily chosen Pquench=10 bar profile rather than a maintained transport equilibrium. This matters because the paper's central claim — that transport, not chemistry, sets the limb distribution of long-lived species — is read directly off this transient tracer field. The citation to Komacek et al. (2019) for insensitivity to the initial tracer profile does not cover this setup: Komacek et al. studied tracers with finite chemical source/sink terms, not a conserved tracer in a regime where tau_chem exceeds 10^19 s, and no sensitivity run is reported here. The issue is addressable, so it warrants a CONDITIONAL verdict rather than rejection.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents 3D GCM simulations of the temperate sub-Neptune K2-18b with the Met Office Unified Model, using a passive tracer to diagnose transport in four rotation states: synchronous, 2:1, 6:1, and 10:1 spin-orbit resonances. After fixed-abundance spin-up runs, the authors enable chemical kinetics and advect a conserved passive tracer with the initial profile given by Eq. (4). They report a detached convective zone at 1--5 bar, an equatorial superrotating jet in all runs, warmer evening terminators, roughly 20% higher tracer mass mixing ratios at the evening terminator, and rotation-dependent latitudinal tracer distributions. They also derive equivalent 1D Kzz profiles using the flux-gradient relation (Eq. 8) and conclude that transport, not chemistry, sets the 3D distribution of long-lived species in the upper atmosphere, with implications for limb-resolved JWST spectra.","tokens_in":31123,"tokens_out":7519,"duration_ms":86301,"significance":"If the central claims hold, the paper provides an important step beyond 1D equilibrium-chemistry models for temperate sub-Neptunes, and its proposed mechanism for evening/morning terminator asymmetries is directly relevant to interpreting limb-resolved transmission spectra. The study's strengths include four self-consistent rotation states, a full correlated-k radiative transfer scheme, explicit momentum-budget and tracer-budget diagnostics, and an explicit caveat that active chemistry is deferred to Part II. The paper also gives a concrete Kzz(P) parameterization that 1D models could adopt. However, the passive-tracer design carries a load-bearing assumption that is not tested, so the quantitative transport claims are not yet fully established.","major_comments":[{"comment":"The passive tracer is initialized with Eq. (4) and evolves with dq/dt = 0, with no source, sink, or reservoir flux. For a conserved tracer in a closed domain, the only steady solution toward which the system can evolve is a globally uniform mass mixing ratio. The \"quasi-steady state\" invoked in §2.3 is therefore a transient set by the initial vertical gradient and the vertical mixing timescale. Figure B2(f) shows the tracer at 0.001 bar evolving through the 5100-day kinetics run, but no plateau criterion is quantified. Consequently, the ~20% evening/morning terminator contrast (Fig. 11) and the equivalent 1D Kzz profile (Fig. 8, Eq. 8) may measure the erosion of the arbitrarily chosen Pquench = 10 bar profile rather than a maintained transport equilibrium. The text in §2.3 also states that the setup represents a species \"sourced uniformly from the deep atmosphere,\" but no source term is present; the deep reservoir is finite. This is load-bearing because the paper's central claim that transport sets the limb distribution of long-lived species is read directly off this tracer field. I recommend adding at least one sensitivity run with a different Pquench or initial slope, and reporting the time evolution of the terminator contrast and of the tracer profile against a quantitative convergence criterion.","section":"§2.3 (Eq. 4), §3.3 (Figs. 8 and 11)"},{"comment":"The statement that the analysis and Kzz are insensitive to the initial tracer profile, citing Komacek et al. (2019), is not supported by the manuscript. Komacek et al. studied tracers with finite chemical source/sink terms and chemical timescales up to 10^6 s, whereas the present simulations use a conserved tracer with no source/sink in a regime where tau_chem exceeds 10^19 s above 1 bar, as the paper itself notes. A citation to a different regime does not establish insensitivity here, and no sensitivity run is shown. This matters because the quantitative claims in §3.3 depend on the initial gradient remaining representative of a real long-lived species with a deep source. Please either provide the missing sensitivity test or soften the claim to state explicitly that the results are provisional on the chosen initial profile.","section":"§2.3 (insensitivity claim)"},{"comment":"The treatment of negative Kzz values by taking the absolute value of Eq. (8) is not adequately justified. Negative flux-gradient ratios are not numerical noise; they indicate genuinely non-diffusive transport, such as the 10:1 SOR high-latitude case where tracer-rich regions coincide with downwelling. Taking absolute values can convert a downgradient-flux violation into an artificially large positive Kzz. Please quantify how often Kzz is negative, show the sensitivity of the fitted Kzz(P) = 3e4 P^-0.61 cm2/s parameterization to including versus excluding or sign-flipping those points, and discuss what the negative values imply for the validity of a 1D diffusive description in those layers.","section":"§3.3 (Eq. 8)"}],"minor_comments":[{"comment":"In Eq. (E5), the first term in the vertical bracket is written as partial([rho w][q] r^2)/partial t; it should be partial r, otherwise the equation is dimensionally inconsistent.","section":"Appendix E, Eq. (E5)"},{"comment":"Section 2.2 states that the final 330 days of the kinetics runs are averaged for analysis, while Section 3.3 says the Kzz profiles are averaged over the last 4600 days. Please clarify which averaging window is used for each result and why the two windows differ.","section":"§2.2 and §3.3"},{"comment":"The caption of Fig. 6(c) gives units of m s^-2 for the contour lines of rho v*/[rho], which is a velocity, and the caption of Fig. 7 refers to \"the black line in panel (a)\" where the summed acceleration appears in panel (b). Please correct these labels.","section":"Fig. 6 and Fig. 7 captions"},{"comment":"The notation \"overbars (primes) and brackets (asterisks)\" is easy to misread; please write explicitly that overbars denote temporal means, primes temporal deviations, brackets zonal means, and asterisks zonal deviations. Also, \"A detailed deviation\" in Appendix D should read \"derivation.\"","section":"§2.3 and Appendix D"},{"comment":"Panel (f) of Fig. B2 is labeled \"Mole fraction at 1 mbar\" but the passive tracer is a mass mixing ratio; the label should be changed to avoid confusion.","section":"Fig. B2"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about the conserved passive tracer lands: the paper's quantitative transport claims are read off a tracer that has no maintained source and whose quasi-steady state is not demonstrated. This is fixable with sensitivity runs or a source-term formulation, so I recommend major revision rather than rejection. The momentum-budget analysis and the four rotation states are careful and will be valuable once the tracer interpretation is placed on firmer footing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is the first to run a 3D GCM with a passive tracer for K2-18b, exploring synchronous rotation and three spin-orbit resonances. The best new thing is the 10:1 SOR result: transient eddies at high latitudes transport tracer upward even where the mean circulation is downwelling. That is a real dynamical finding and worth having out there. The broader circulation story is also well argued. The detached convective zone between 1 and 5 bar, driven by CO2 and CH4 absorption, is plausible; the momentum budget analysis for the superrotating jet is careful; the four rotation runs are mutually consistent. The paper is honest that this part is dynamics only, with chemistry to follow in Part II.\n\nNow the soft spots. The main one is load-bearing: the passive tracer has no source or sink (dq/dt = 0) and is initialized with an arbitrary steep vertical gradient breaking at 10 bar. In a closed domain, the only steady state for such a tracer is a uniform mixing ratio. So the tracer structure at the end of 5100 days is a transient, set by the initial gradient and the mixing timescale. The paper calls it quasi-steady and asserts insensitivity to the initial profile, citing Komacek et al. (2019), but that study used tracers with finite chemical source/sink terms. No sensitivity run is shown, and Figure B2(f) does not demonstrate a plateau. The 20% terminator asymmetry and the equivalent Kzz profile are read directly off this tracer field. They may be right in a qualitative sense, but as quantitative numbers they are not yet supported. This needs fixing, either with a tracer that has a deep source and upper sink, or with a demonstration that the results converge for different initial profiles and longer runs.\n\nTwo secondary concerns. The model metallicity and initial Kzz are tuned to the same JWST spectra that Part II will interpret; that is calibration, not circularity, but it is worth remembering when the second paper makes observational claims. And the recent Schmidt et al. (2025) reanalysis casts doubt on the CO2 detection the setup relies on; the authors acknowledge it, which is good.\n\nWho is this for? Exoplanet atmospheric modelers, especially anyone building 1D chemical models with Kzz for temperate sub-Neptunes, and anyone trying to interpret JWST limb asymmetries. It deserves a serious referee, but the referee should push for the tracer sensitivity test before accepting the quantitative claims.\n\nRecommendation: send to peer review. It is a solid dynamics paper with one unverified assumption that is explicitly checkable.","headline":"First 3D transport picture for a temperate sub-Neptune, with a genuinely new high-latitude eddy result in the 10:1 SOR case, but the quantitative tracer claims rest on an unverified conserved-tracer assumption.","tokens_in":31681,"tokens_out":3657,"would_cite":true,"duration_ms":39882,"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":"On the temperate sub-Neptune K2-18b, atmospheric transport, not chemistry, sets the three-dimensional distribution of long-lived molecules, with an equatorial jet producing a roughly 20% morning–evening abundance asymmetry.","keywords":["sub-Neptune","K2-18b","atmospheric circulation","passive tracer transport","superrotation","limb asymmetry","vertical mixing","general circulation model"],"falsifier":"A decisive test would be a limb-resolved transmission spectrum of K2-18b comparing the evening and morning terminators: if the evening terminator is not systematically warmer and richer in methane and carbon dioxide (on the order of 20% column difference), the transport-asymmetry claim fails. A second, model-internal test would be to rerun the fastest 10:1 spin-orbit simulation with a different initial tracer profile (e.g., a shallower gradient or a deep source at a different pressure) and check whether the high-latitude transient-eddy transport and the derived K_zz remain unchanged.","tokens_in":30545,"feed_emoji":"🪐","tokens_out":5826,"duration_ms":56650,"temperature":0.7,"pith_summary":"On the temperate sub-Neptune K2-18b, the paper argues, winds and eddies—not chemical reactions—set the three-dimensional distribution of long-lived molecules in the upper atmosphere. Using a general circulation model with a passive tracer to stand in for species with very long chemical lifetimes, it finds that an equatorial eastward jet pushes material toward the evening terminator, producing a roughly 20% higher tracer abundance there than at the morning terminator. It also finds a detached convective zone between 1 and 5 bar, caused by strong CO2 and CH4 absorption, which drives vigorous vertical mixing. If true, this means one-dimensional equilibrium chemistry models cannot reliably predict the limb-resolved spectra that JWST observes, and rotation state—synchronous versus asynchronous—strongly alters where molecules accumulate.","feed_headline":"K2-18b's upper air is shaped by winds, not chemistry","feed_subtitle":"3D model finds a 20% evening-vs-morning abundance gap, so 1D chemical models can't predict JWST limb spectra.","key_machinery":"The central object is a passive tracer: a mass-mixing-ratio field q that obeys dq/dt = 0, advected by the model's resolved winds with no source, sink, or radiative feedback, initialized with a steep vertical gradient (q = 1e-5 below 10 bar, declining as (P/10 bar)^1.5 above). It is used as a diagnostic for long-lived chemical species, and the paper diagnoses transport by decomposing the zonal- and time-mean tracer budget into mean-flow, stationary-eddy, and transient-eddy terms, and by estimating an equivalent 1D vertical eddy diffusivity K_zz from the flux-gradient relation K_zz = -<rho q w>/<rho dq/dr>. The K_zz profile and the eddy-mean decomposition together carry the argument: they convert a 3D circulation into a form 1D models and spectral interpretations can use.","core_discovery":"The paper's central claim is that transport dominates composition on temperate sub-Neptunes: for K2-18b, where chemical timescales exceed $10^{10}$ seconds above roughly 10 bar, a passively advected tracer representing long-lived species is redistributed by atmospheric circulation faster than chemistry can act. In all simulated rotation states (synchronous, 2:1, 6:1, and 10:1 spin-orbit resonances), an equatorial superrotating jet carries tracer eastward, making the evening terminator both warmer and about 20% richer in tracer than the morning terminator. Rotation has little effect on the global-mean vertical mixing strength, yielding a universal equivalent eddy-diffusion profile that decays with pressure and is enhanced inside the 1–5 bar detached convective zone. But rotation strongly controls the latitudinal pattern: in slow-rotating cases tracer is concentrated in low- and mid-latitude upwelling branches, while in the fastest 10:1 case transient eddies at latitudes above 70 degrees lift tracer from depth even where the mean circulation is downwelling. The paper concludes that 1D models can still capture global-mean vertical structure, but any interpretation of limb-resolved or latitude-dependent spectra must account for the three-dimensional transport patterns.","pith_inferences":["We infer that the same transport mechanism should operate on other temperate sub-Neptunes with similar equilibrium temperatures and orbital periods, so the morning-evening asymmetry may be a general feature rather than unique to K2-18b.","A testable extension would be to compute synthetic transmission spectra from the 3D tracer (and later chemical) fields for both terminators separately and compare directly with time-resolved JWST transits, which would isolate the asymmetry without waiting for new observations.","If the 20% asymmetry survives in Part II's active-chemistry runs, it implies that retrieved CH4 and CO2 abundances from JWST are systematically offset depending on which portion of the transit is fitted, possibly explaining some of the current scatter among retrieval results."],"forward_implications":["If correct, 1D forward models of temperate sub-Neptunes should be interpreted as representing global-mean vertical structure only; they cannot reproduce limb-dependent abundance patterns.","Spectra that average over both terminators will mix different compositions, so retrieving abundances without accounting for the ~20% evening/morning asymmetry will bias the retrieved metallicity and C/O ratio.","For asynchronously rotating planets (fast spin-orbit resonances), high-latitude transport can create polar or high-latitude chemical enrichment, which would appear in spectra only if the planet is observed at favorable geometries.","The detached convective zone between 1 and 5 bar, driven by CO2 and CH4 absorption, implies that vertical mixing peaks at intermediate depths, and simple K_zz prescriptions should be replaced by pressure-dependent profiles like K_zz = 3e4 * (1/P_bar)^0.61 cm^2/s above 1 bar.","The derived equivalent K_zz can be ported directly into 1D chemical kinetics models, giving a physically grounded mixing profile for a planet that orbits an M star."],"supporting_citations":[{"why":"Supplies the basis for the claim that the derived K_zz and tracer statistics are insensitive to the initial tracer profile when the vertical gradient is steep and the simulation reaches quasi-steady state.","marker":"Komacek et al. 2019"},{"why":"Provides the flux-gradient relationship (their equation 23) used to estimate the equivalent 1D vertical eddy diffusivity from the 3D tracer fields.","marker":"Parmentier et al. 2013"},{"why":"Establishes the tracer-transport diagnostic framework and the interpretation of tracer–vertical-velocity correlations used to explain tracer accumulation in upwelling regions.","marker":"Zhang & Showman 2018a"},{"why":"Provides the counterpart framework for eddy-mean tracer decomposition and the treatment of non-diffusive 3D transport when estimating K_zz.","marker":"Zhang & Showman 2018b"},{"why":"Gives the mechanism of equatorial superrotation by stationary eddy momentum convergence, which the paper identifies as the driver of the eastward jet that creates the terminator asymmetry.","marker":"Showman & Polvani 2011"},{"why":"Provides the Arrhenius-like chemical timescale fits used to compare tau_chem with mixing and advection timescales, establishing that transport dominates above ~10 bar.","marker":"Zahnle & Marley 2014"},{"why":"Provides the JWST-observed CH4 and CO2 abundances that define the atmospheric composition and metallicity used to set up the simulations and motivate the deep-source tracer scenario.","marker":"Madhusudhan et al. 2023"}],"fun_headline_variants":["Winds, not chemistry, set K2-18b's air composition","K2-18b's evening air is 20% richer—winds, not chemistry","Transport, not chemistry, shapes K2-18b's atmosphere","Winds drive K2-18b's chemistry—evening vs morning gap","On K2-18b, winds beat chemistry for air composition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a conserved tracer with an arbitrarily chosen initial depth profile and no source terms faithfully represents real long-lived chemical species, so that the derived transport statistics and the 20% terminator asymmetry are not artifacts of the tracer's initial condition or the run length.","fun_headline_variants_meta":{"raw":{"variants":["Winds, not chemistry, set K2-18b's air composition","K2-18b's evening air is 20% richer—winds, not chemistry","Transport, not chemistry, shapes K2-18b's atmosphere","Winds drive K2-18b's chemistry—evening vs morning gap","On K2-18b, winds beat chemistry for air composition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000699,"raw_usage":{"total_tokens":3268,"prompt_tokens":1164,"completion_tokens":2104,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":780,"completion_tokens_details":{"reasoning_tokens":2002}},"tokens_in":780,"tokens_out":2104,"duration_ms":15487,"temperature":1.0,"reasoning_tokens":2002,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:29:52.219398+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be a limb-resolved transmission spectrum of K2-18b comparing the evening and morning terminators: if the evening terminator is not systematically warmer and richer in methane and carbon dioxide (on the order of 20% column difference), the transport-asymmetry claim fails. A second, model-internal test would be to rerun the fastest 10:1 spin-orbit simulation with a different initial tracer profile (e.g., a shallower gradient or a deep source at a different pressure) and check whether the high-latitude transient-eddy transport and the derived K_zz remain unchanged.","supporting_citations":[],"review_version":1}