{"id":"e88e42d4-c8c0-4836-a55f-b0352e28b9f3","arxiv_id":"2607.08026","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A calibrated thick-disk semianalytic model plus MCMC retrieval quantifies SED constraints on CPD parameters and recovers consistent masses and accretion rates for PDS 70 b/c and GQ Lup b.","lead":"The authors build a fast semianalytic model of geometrically thick disks around forming giant planets that matches 2D radiative-transfer simulations, then use MCMC fits to quantify which planet and disk parameters spectra can actually constrain. Upcoming telescopes will find many such systems; this framework shows what can be measured and applies it to PDS 70 and GQ Lup b.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.5","headline":"Sparse 40-model RBF calibration of α_d, q, f_L is the load-bearing soft spot for the claimed posterior widths.","rationale":"The reader correctly isolates the sparse RBF calibration of the three structure parameters as the weakest link supporting the quantitative posterior claims. The synthetic MCMC experiments and real-system applications are internally consistent once those parameters are granted, so the paper still warrants CONDITIONAL rather than REJECT; the concern only reinforces that the acceptance must remain conditional on the calibration remaining accurate across the full retrieval volume. No stronger internal inconsistency appears: energy conservation, opacity power-law caveats, and fixed R_p/λ_d assumptions are already flagged by the authors and do not overturn the luminosity-first degeneracy structure. The concrete hold-out test above would settle whether the 40-model grid is adequate or whether denser training (or a more flexible photosphere) is required before the dex-level claims can be treated as robust.","tokens_in":38256,"tokens_out":797,"duration_ms":8277,"concrete_test":"Hold out an independent RAD+ grid of ≥20 models concentrated in the mature-super-Jupiter and high-Ṁ corners (optically thin outer disk and thick-envelope regimes). For each, (i) interpolate α_d, q, f_L from the original 40-point training set, (ii) generate the SAM SED, and (iii) run the identical MCMC pipeline used for Figs. 7–9. If recovered L_tot or τ_acc widths exceed the paper’s quoted 0.05–0.8 dex by >50 %, or if true values fall outside the 2σ posterior more than ~5 % of the time, the claimed constraint precisions are overstated.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The strongest claim (L_tot ≲0.1 dex; FIR-driven τ_acc ~0.35–0.8 dex for optically thick disks; NIR+MIR τ_env for embedded cases) rests on the SAM matching RAD+ SEDs after structure parameters α_d, q, f_L are supplied by multiquadric RBF interpolation on a 40-point RAD+ training grid (Sec. 2.2, Appendix A). Those three parameters fully set the constant-aspect-ratio photosphere, T(R) power law, and self-shadowing factor f_L that enter the energy-conserving T_X and envelope T_C solutions (Eqs. 7–8, 15–16, A.1). The paper reports only an average ~20 % RMS SED error on 40 random test models (Fig. 4) and that true parameters lie inside 1σ of the MCMC posteriors for those tests. That is necessary but not sufficient: the claimed dex-level posterior widths can be systematically optimistic if the RBF surface is under-sampled in the high-M_p / low-Ṁ or high-Ṁ / large-a corners that control optical-depth transitions and FIR outer-disk emission (the exact regimes that distinguish the three fiducials in Secs. 3.1–3.3). The constant-aspect-ratio ansatz itself is also untested against RAD+ surfaces that may flare or warp. If either the interpolation or the ansatz fails locally, the FIR τ_acc and NIR+MIR τ_env constraints become unreliable even while broadband SED shapes still look acceptable.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper constructs a calibrated semianalytic model (SAM) for the continuum SEDs of geometrically thick circumplanetary disks (CPDs) and their host protoplanets, then uses MCMC retrievals on synthetic RAD+ SEDs and on real photometry of PDS 70 b/c and GQ Lup b to quantify which system parameters can be constrained. Structure parameters α_d, q and f_L are interpolated from a 40-model RAD+ training grid so that the SAM matches RAD+ SEDs to ~20 % RMS (Fig. 4). For optically thick disks the full SED (especially FIR) is claimed to constrain total luminosity to ≲0.1 dex and the accretion timescale to ~0.35–0.8 dex; for embedded systems NIR+MIR constrain envelope optical depth. Applications to PDS 70 and GQ Lup b recover luminosities tightly and masses/accretion rates consistent with independent estimates once modest extinction is allowed.","tokens_in":38796,"tokens_out":1178,"duration_ms":10742,"significance":"If the claimed posterior widths hold under the stated assumptions, the work supplies a practical, computationally cheap retrieval framework for the next generation of unresolved CPD detections (JWST, ELT/METIS, and any future FIR capability). The explicit mapping of wavelength bands onto derived quantities (L_tot, τ_acc, M_dust, τ_env) and the demonstration that FIR continuum can break the M_p–Ṁ degeneracy for optically thick CPDs are useful, falsifiable predictions. The real-system applications already give concrete numbers for PDS 70 and GQ Lup b and motivate MIR follow-up. Strengths include the direct RAD+ validation set, energy-conserving construction of T_X and T_C, and transparent discussion of model limitations (opacity, fixed R_p/λ_d, continuum-only).","major_comments":[{"comment":"Sec. 2.2 and Appendix A: the three structure parameters (α_d, q, f_L) that fully set the constant-aspect-ratio photosphere, T(R) and self-shadowing are obtained by multiquadric RBF interpolation on only 40 RAD+ training models. The paper reports only a global ~20 % RMS SED error on 40 random test models (Fig. 4) and that injected parameters lie inside 1σ. That is necessary but not sufficient for the dex-level posterior widths claimed in Secs. 3.1–3.3 (Figs. 7–9). The FIR τ_acc and NIR+MIR τ_env constraints are controlled by the high-M_p/low-Ṁ and high-Ṁ/large-a corners where optical-depth transitions and outer-disk emission change. Without leave-one-out or denser-grid tests that quantify local interpolation error in those corners, the reported posterior widths may be systematically optimistic. A short appendix quantifying RBF residuals versus M_p, Ṁ and a (or a denser training set) is ne","section":null},{"comment":"Sec. 2.2 and Eqs. (7)–(8), (15)–(16): the constant-aspect-ratio conical photosphere and the step-function cutoff ψ_c used for energy conservation are strong geometric assumptions. The paper never shows that RAD+ τ=1 surfaces are well approximated by a single α_d, nor how much flaring/warping residual remains after the median α_d is taken. Because the FIR outer-disk emission and the self-shadowing factor f_L both depend on this geometry, residual mismatch can bias the very quantities (q, R_C temperature, τ_acc) that the paper uses to claim that FIR breaks the M_p–Ṁ degeneracy. A direct comparison of SAM versus RAD+ surface shapes (or of the resulting FIR SEDs when α_d is forced to the RAD+ median) would make this load-bearing step transparent.","section":null},{"comment":"Sec. 3 (opening paragraphs) and Table 2: R_p and λ_d are fixed a priori while the text acknowledges that plausible ranges (ΔR_p ~0.5 dex, λ_d down to 0.3) shift L_tot and the outer radius. The synthetic-retrieval experiments that produce the headline 0.05–0.15 dex L_tot and 0.35–0.8 dex τ_acc widths never re-run with these parameters free or marginalized. Because the real-system applications (Sec. 4) also fix R_p=2 R_J, the quoted uncertainties on M_p and Ṁ for PDS 70 and GQ Lup b are conditional on that choice. At minimum the paper should show one set of synthetic posteriors with R_p (and preferably λ_d) free so that readers can judge how much the claimed precisions degrade.","section":null}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The useful part of this paper is the quantitative forecast of what continuum SEDs actually constrain for geometrically thick CPDs. They build a fast SAM calibrated to their RAD+ runs, show ~20% RMS SED agreement on a 40-model hold-out set, recover injected parameters inside 1σ, and then map NIR/MIR/FIR onto L_tot ≲ 0.1 dex, FIR-driven τ_acc for optically thick disks, and NIR+MIR τ_env for embedded cases. The PDS 70 and GQ Lup b applications are consistent with dynamical and Hα numbers once modest extinction is allowed, and the entropy upper limits are a clean extra product.\n\nWhat is new is the thick-disk geometry (constant-aspect-ratio photosphere, inner wall, self-shadowing f_L, three-region envelope) plus the wavelength-by-wavelength MCMC forecast. Their earlier thin-disk SAMs did not match RAD+ well enough for this; this one does under the stated assumptions. The math is standard energy conservation and plane-parallel slabs; the citation pattern is normal self-citation of the RAD+ series plus the usual CPD literature. No circularity: structure parameters are fixed from an independent grid before the retrievals run.\n\nThe soft spot the stress-test flags is real but proportionate. α_d, q, and f_L come from multiquadric RBF on only 40 training models. That is sparse for the high-M_p/low-Ṁ and high-Ṁ/large-a corners that control the optical-depth transition and FIR outer-disk emission. The constant-aspect-ratio ansatz is also untested against flaring or warping. If the interpolant or the ansatz fails locally, the claimed dex-level posterior widths on τ_acc and τ_env can be optimistic even while broadband shapes still look fine. Fixed R_p, λ_d, and power-law opacity (missing the 10 µm feature and the long-wavelength drop) are secondary and already flagged by the authors. No public code is a practical annoyance, not a scientific flaw.\n\nThis is for people who will actually fit JWST/ELT CPD SEDs or design FIR follow-up. It deserves a serious referee. I would cite the degeneracy maps and the PDS 70 numbers. Engage; the calibration grid is the one place that needs pressure.","headline":"Usable thick-disk CPD retrieval that cleanly maps wavelength bands onto L_tot, τ_acc, and τ_env, with the sparse RBF calibration as a real but not fatal soft spot.","tokens_in":39313,"tokens_out":637,"would_cite":true,"duration_ms":7817,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"A fast thick-disk model turns continuum SEDs of forming giant planets into quantitative constraints on luminosity, accretion timescale, and envelope extinction.","keywords":["circumplanetary disks","planet formation","spectral energy distributions","parameter retrieval","MCMC fitting","protoplanets","PDS 70","GQ Lup b"],"falsifier":"Obtain simultaneous high-resolution NIR+MIR+FIR photometry of a known CPD system (or a high-fidelity RAD+ synthetic SED outside the training set) and check whether the SAM posterior recovers the true Mp and Ṁ within the claimed 0.35–0.8 dex; systematic failure outside those bounds would falsify the calibration.","tokens_in":39175,"feed_emoji":"🌌","tokens_out":689,"duration_ms":6935,"temperature":0.7,"pith_summary":"Giant planets form while still surrounded by thick disks of gas and dust drawn from the larger circumstellar disk. Because even the next generation of telescopes cannot spatially resolve these systems, their spectral energy distributions are the only practical observables. This paper builds a calibrated semianalytic model of geometrically thick circumplanetary disks that reproduces the SEDs of expensive two-dimensional radiative-transfer calculations, then uses Markov-chain Monte Carlo fits to synthetic and real data to measure how tightly those SEDs constrain the underlying parameters. Multiband continuum photometry can pin down the total system luminosity to better than 0.1 dex; for optically thick disks the far-infrared further separates planet mass from accretion rate, while near- plus mid-infrared slopes diagnose envelope optical depth when the planet is still embedded. Applied to PDS 70 b/c and GQ Lup b, the same machinery recovers luminosities, rough masses and accretion rates, and upper limits on local extinction that are consistent with independent dynamical and H-alpha estimates. The result is a practical retrieval tool that turns future ELT and JWST detections into quantitative statements about how and how fast giant planets grow.","feed_headline":"Thick-disk SEDs pin down forming-planet luminosity to 0.1 dex","feed_subtitle":"Far-IR further separates mass from accretion rate; NIR+MIR diagnose envelope extinction","key_machinery":"The thick-disk semianalytic model (SAM): three structure parameters (α_d, q, f_L) interpolated from a RAD+ training grid fully determine the disk photosphere, self-shadowing, and emergent SEDs, enabling 10^8-fold speed-up over full radiative-transfer calculations so that MCMC retrieval becomes practical.","core_discovery":"A calibrated semianalytic model of geometrically thick circumplanetary disks, when fit by MCMC to continuum SEDs, recovers the total system luminosity to ≲0.1 dex and, for optically thick disks, the accretion timescale (hence Mp and Ṁ separately) to roughly 0.35–0.8 dex, while joint near- and mid-infrared data constrain the line-of-sight envelope optical depth of embedded systems.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Thick CPD SEDs recover system luminosity to 0.1 dex via MCMC","Semianalytic model pins protoplanet luminosity to ≲0.1 dex","SED fits of thick disks constrain luminosity and accretion timescale","MCMC on CPD continuum SEDs yields 0.1 dex luminosity precision","Joint IR data separate mass, accretion rate, and envelope extinction"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"The three structure parameters that set the entire disk temperature and self-shadowing are assumed to remain accurate when interpolated from a sparse grid of only forty numerical training models.","fun_headline_variants_meta":{"raw":{"variants":["Thick CPD SEDs recover system luminosity to 0.1 dex via MCMC","Semianalytic model pins protoplanet luminosity to ≲0.1 dex","SED fits of thick disks constrain luminosity and accretion timescale","MCMC on CPD continuum SEDs yields 0.1 dex luminosity precision","Joint IR data separate mass, accretion rate, and envelope extinction"]},"model":"grok-4.5","effort":"low","cost_usd":0.003642,"raw_usage":{"total_tokens":1129,"prompt_tokens":735,"num_sources_used":0,"completion_tokens":98,"cost_in_usd_ticks":36420000,"prompt_tokens_details":{"text_tokens":735,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":296,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":735,"tokens_out":98,"duration_ms":3881,"temperature":1.0,"reasoning_tokens":296,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-10T13:28:18.819825+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Obtain simultaneous high-resolution NIR+MIR+FIR photometry of a known CPD system (or a high-fidelity RAD+ synthetic SED outside the training set) and check whether the SAM posterior recovers the true Mp and Ṁ within the claimed 0.35–0.8 dex; systematic failure outside those bounds would falsify the calibration.","supporting_citations":[],"review_version":1}