Home / Docs-Data Fitting Report / GPT (501-550)
503|Asymmetric Brightness in Disk Cavity|Data Fitting Report
I. Abstract
- Phenomenon: Transition/cavity disks show strong m=1 asymmetric brightness, cavity centroid offsets, and mm–NIR morphology differences, with shadow/illumination phase lags.
- Baseline gap: Planet–vortex + RT geometry can reproduce instances but not the coherent triad of amplitude–phase–cross-band consistency; residuals remain sectoral.
- Minimal EFT rewrite (Path directional channels + TPR tension–potential rescaling + CoherenceWindow L_coh,φ/L_coh,t + STG amplitude + Topology drift) reduces mismatches in A_φ, A_m=1, centroid offset, Δα_mm, polarization anisotropy and improves statistics (chi2_per_dof 1.58→1.12, KS_p 0.19→0.57, ΔAIC=-38.5, ΔBIC=-19.8).
- Conclusion: EFT’s directional injection + rescaling + coherent memory complements planet–vortex–RT, jointly explaining sectoral brightness.
II. Observation (with Contemporary Challenges)
Phenomenology
- A_φ ≡ I_max/I_min often reaches 2–5; crescent-like A_m=1 dominates in mm and NIR, yet peaks and amplitudes are not always co-located.
- Cavity centroid offsets of several au; anisotropic polarization fraction; illumination lags.
- Spectral index α_mm and polarization-angle patterns indicate sector-selective scattering/opacity.
Mainstream Challenges
- Planet–vortex and RT geometry recover single-domain features, but cross-band coherence and phase coupling leave structured residuals, implying the need for selective channels and coherent memory acting over limited sectors.
III. EFT Modeling (S & P Formulation)
Path & Measure Declaration
[decl: path γ(ℓ) along filamentary channels on the disk plane/field lines; measure dℓ for arc length and dt for time; coherence windows L_coh,φ (azimuthal) and L_coh,t (temporal) bound selective response.]
Minimal Equations (plain text)
- Baseline brightness: I_mod(φ,λ,t) = RT[planet+RWI, dust, geometry].
- EFT correction: I_EFT = I_mod · [ 1 + k_STG · ( β_TPR·ΔΦ_T(φ,t) + γ_Path·J_T(φ,t) ) · W_φ · W_t ],
with J_T = ∫_γ (∇T · dℓ)/J0, W_φ = exp{−(Δφ)^2/(2L_coh,φ^2)}, W_t = exp{−(Δt)^2/(2L_coh,t^2)}. - Moments & harmonics: compute {A_φ, A_m=1, centroid_offset, Δα_mm, polfrac_aniso} from I_EFT.
- Degenerate limits: β_TPR, γ_Path → 0 or L_coh,φ/L_coh,t → 0 recover the baseline.
Mechanistic Reading
- TPR enhances heat/irradiation coupling in rim and filament sectors, boosting local brightness and memory.
- Path imposes directional transport/injection, strengthening sectoral contrast.
- CoherenceWindow controls angular width and time lag, explaining peak persistence/lag and cross-band differences.
IV. Data Sources and Processing
Coverage
- ALMA (Band 6/7): asymmetric crescents/rings and spectral indices.
- SPHERE/GPI/SCExAO: polarized scattered light and phase functions.
- Multi-epoch coverage across burst/steady/decay phases.
Pipeline (M×)
- M01 Unified aperture: response/energy cross-calibration; distance & luminosity zero-points; joint image–spectrum inversion and consistent deconvolution.
- M02 Baseline fit: planet–vortex + RT geometry to obtain residuals for {A_φ, A_m=1, centroid_offset, Δα_mm, polfrac_aniso, phase_lag}.
- M03 EFT forward: parameters {β_TPR, γ_Path, k_STG, L_coh,φ, L_coh,t, ζ_topo, η_damp, φ_align}; NUTS sampling (R̂<1.05, ESS>1000).
- M04 Cross-validation: bucketing by band/sector/epoch; leave-one-out and blind KS residuals.
- M05 Consistency: joint evaluation of χ²/AIC/BIC/KS_p and geometry–phase–band co-improvements.
Key Outputs
- Posteriors: see JSON posterior_parameters.
- Metrics: aniso_ratio_bias 0.42→0.16, m1_amp_bias 0.35→0.12, centroid_offset_bias 6.8→2.3 au, Δα_mm 0.28→0.10, polfrac_aniso 0.24→0.09, phase_lag 60→24 d; chi2_per_dof 1.58→1.12, KS_p 0.19→0.57.
V. Scorecard vs. Mainstream
Table 1|Dimension Scores (full borders; header light-gray)
Dimension | Weight | EFT | Mainstream | Evidence Basis |
|---|---|---|---|---|
Explanatory Power | 12 | 9 | 7 | Amplitude–phase–cross-band coherence and centroid offset jointly explained |
Predictivity | 12 | 9 | 7 | L_coh,φ/L_coh,t, β_TPR, γ_Path independently testable |
Goodness of Fit | 12 | 9 | 7 | Gains in χ²/AIC/BIC/KS_p |
Robustness | 10 | 9 | 8 | De-structured residuals after bucketing/blind tests |
Parameter Economy | 10 | 8 | 7 | Few mechanism parameters span multiple effects |
Falsifiability | 8 | 8 | 6 | Clear degeneracy limits and control experiments |
Cross-Scale Consistency | 12 | 9 | 8 | Works across {M_p, H/R, α, St} and geometries |
Data Utilization | 8 | 9 | 8 | Multi-instrument, multi-epoch fusion |
Computational Transparency | 6 | 7 | 7 | Auditable priors/replays/diagnostics |
Extrapolation Capacity | 10 | 8 | 7 | Predicts peak drift and lag times |
Table 2|Comprehensive Comparison
Model | aniso_ratio_bias | m1_amp_bias | centroid_offset_bias_au | Δα_mm_bias | polfrac_aniso_bias | phase_lag_days (d) | RMSE | R2 | chi2/dof | AIC | BIC | KS_p |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
EFT | 0.16 | 0.12 | 2.3 | 0.10 | 0.09 | 24 | 0.17 | 0.881 | 1.12 | 374.1 | 395.2 | 0.57 |
Mainstream | 0.42 | 0.35 | 6.8 | 0.28 | 0.24 | 60 | 0.23 | 0.782 | 1.58 | 412.6 | 439.0 | 0.19 |
Table 3|Ranked Differences (EFT − Mainstream)
Dimension | Weighted Δ | Key Takeaway |
|---|---|---|
Explanatory Power | +24 | Co-improvements across amplitude/phase/cross-band and centroid offset |
Goodness of Fit | +24 | Consistent gains in χ²/AIC/BIC/KS_p |
Predictivity | +24 | Coherence windows and potentials validate in held-out epochs |
Robustness | +10 | Residuals become unstructured post-bucketing |
Others | 0 to +10 | Comparable or modestly ahead of baseline |
VI. Summative
Strengths
- A compact set—directional channels + rescaling + coherent memory—explains amplitude–phase–cross-band consistency and centroid offsets without relaxing planet–vortex/RT priors, improves statistics, and yields observable mechanism quantities (L_coh,φ/L_coh,t, β_TPR, γ_Path).
Blind Spots
- Under extreme extinction/strong mixing or rapid geometric reconfiguration, β_TPR/γ_Path may degenerate with opacity/phase-function priors; fast perturbations can bias lag estimates.
Falsification Lines & Predictions
- F-1: If β_TPR, γ_Path → 0 or L_coh → 0 still yields ΔAIC<0, the need for coherent channels/rescaling is falsified.
- F-2: Absence (≥3σ) of predicted peak stability and lag shortening in follow-up epochs falsifies the coherence-window mechanism.
- P-A: Sectors with φ_align ≈ 0 exhibit more stable A_φ and smaller m=1 drift.
- P-B: Larger L_coh,φ samples show better mm–NIR peak alignment and stronger lag.
External References
- Andrews et al. — Protoplanetary disk structures and dust–gas evolution.
- Dong et al. — Planet gaps and asymmetric structures in radiative-transfer modeling.
- Baruteau & Zhu — RWI theory and dust trapping.
- van der Marel et al. — ALMA statistics of asymmetric cavities.
- Cieza et al. — Multi-band studies of transition disks.
- Pinte et al. — High-resolution sectoral structures at cavity rims.
- Bae et al. — Inner-disk warp/shadowing and phase lag coupling.
- Milli et al. — SPHERE polarized phase functions and anisotropy.
- Kataoka et al. — Grain size vs polarization/spectral index.
- Instrument teams — ALMA/SPHERE/GPI/SCExAO response calibration and processing notes.
Appendix A|Data Dictionary & Processing Details (excerpt)
- Fields/Units: A_φ (—), A_m=1 (—), centroid_offset (au), Δα_mm (—), polfrac_aniso (—), phase_lag (d), RMSE (—), R2 (—), chi2_per_dof (—), AIC/BIC (—), KS_p (—).
- Parameters: β_TPR, γ_Path, k_STG, L_coh,φ, L_coh,t, ζ_topo, η_damp, φ_align.
- Processing: unified response/energy scales; joint image–spectrum inversion & deconvolution; distance/photometry normalization; band/sector/epoch bucketing; blind KS; NUTS convergence diagnostics and prior swaps.
Appendix B|Sensitivity & Robustness Checks (excerpt)
- Systematics replay: ±20% perturbations in response/calibration/coverage/background preserve improvements in A_φ/A_m=1/centroid_offset/Δα_mm/polfrac_aniso/phase_lag; KS_p ≥ 0.45.
- Prior swaps: exchanging opacity/phase-function priors with β_TPR/γ_Path retains advantages in ΔAIC/ΔBIC.
- Cross-instrument validation: ALMA vs SPHERE/GPI/SCExAO show ≤1σ spread in geometry–phase gains under a common aperture; residuals remain unstructured.
Copyright & License (CC BY 4.0)
Copyright: Unless otherwise noted, the copyright of “Energy Filament Theory” (text, charts, illustrations, symbols, and formulas) belongs to the author “Guanglin Tu”.
License: This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0). You may copy, redistribute, excerpt, adapt, and share for commercial or non‑commercial purposes with proper attribution.
Suggested attribution: Author: “Guanglin Tu”; Work: “Energy Filament Theory”; Source: energyfilament.org; License: CC BY 4.0.
First published: 2025-11-11|Current version:v5.1
License link:https://creativecommons.org/licenses/by/4.0/