GPT (301-350) | 329 | Long-Term Trend in Lens Time-Delay Drift | Data Fitting Report

Home ／ Docs-Data Fitting Report ／ GPT (301-350)

329 | Long-Term Trend in Lens Time-Delay Drift | Data Fitting Report

{
  "spec_version": "EFT Data Fitting English Report Specification v1.2.1",
  "report_id": "R_20250909_LENS_329_EN",
  "phenomenon_id": "LENS329",
  "phenomenon_name_en": "Long-Term Trend in Lens Time-Delay Drift",
  "scale": "Macro",
  "category": "LENS",
  "language": "en",
  "eft_tags": [
    "TimeDelay",
    "Drift",
    "Path",
    "TensionGradient",
    "CoherenceWindow",
    "ModeCoupling",
    "Topology",
    "Sea Coupling",
    "Recon",
    "Damping",
    "ResponseLimit",
    "Clock"
  ],
  "mainstream_models": [
    "ΛCDM + GR strong-lensing baseline: the time delay `Δt` is set by the Fermat potential; macromodel (EPL/SIE+γ) plus geometric/potential terms are approximately constant over the monitoring span. Source variability is modeled as a stationary stochastic process (DRW/OU). After time-standard and barycentric corrections (UTC/TAI→TDB/TCB; BJD), `Δt` is treated as constant or zero-mean noise.",
    "Supplements: microlensing time delay, slowly varying LOS potential, substructure and host-environment adjustments, seasonal sampling and instrument zero-point drifts, dispersion and bandpass differences creating cross-band delay offsets; MST and shear–ellipticity degeneracy affect multi-season joint fits.",
    "Systematics: PSF/registration/background drifts, mass–light decomposition biases, residual time-standard/barycentric corrections, differential photometry and aperture effects, data gaps and seasonal aliasing, inconsistent uv/image-domain weighting."
  ],
  "datasets_declared": [
    {
      "name": "COSMOGRAIL (optical long-term monitoring; multi-season)",
      "version": "public",
      "n_samples": "~70 lenses / >10^4 nightly epochs"
    },
    {
      "name": "TDCOSMO/H0LiCOW (time-delay lenses; joint dynamics/environment)",
      "version": "public",
      "n_samples": "~15 systems"
    },
    {
      "name": "DES/STRIDES & HSC (long-baseline variability and candidate confirmation)",
      "version": "public",
      "n_samples": "~200 systems"
    },
    {
      "name": "JVLA/ALMA (radio/mm monitoring; cross-band delays)",
      "version": "public",
      "n_samples": "~60 systems"
    },
    {
      "name": "Simulations: macromodel + microlensing + LOS potential multi-season replay (with time-standard/barycentric/zero-point/aperture injections)",
      "version": "public",
      "n_samples": ">10^3 realizations, 3–12 yr baselines"
    }
  ],
  "metrics_declared": [
    "drift_rate_bias (ms/yr; model bias of long-term slope `|dΔt/dt|`)",
    "td_trend_rms (ms/yr; season-level slope RMS after detrending)",
    "microlens_td_resid (ms; microlensing time-delay residual amplitude)",
    "seasonal_phase_resid (deg; seasonal phase/aliasing residual)",
    "structure_func_slope_bias (—; bias of structure-function index β)",
    "clock_bary_bias (ms; residual from time-standard/barycentric corrections)",
    "los_var_resid (—; LOS potential-variation residual amplitude)",
    "crossband_delay_bias (ms; bias in inter-band delay differences)",
    "KS_p_resid",
    "chi2_per_dof",
    "AIC",
    "BIC"
  ],
  "fit_targets": [
    "Under a unified pipeline (PSF/registration/time-standard & barycentric corrections/differential photometry/selection), jointly reduce `drift_rate_bias`, `td_trend_rms`, `microlens_td_resid`, `seasonal_phase_resid`, `structure_func_slope_bias`, and `clock_bary_bias/los_var_resid/crossband_delay_bias`, while increasing `KS_p_resid`.",
    "Do not degrade positions/fluxes/arc geometry or two-point statistics; maintain consistency across seasons/facilities/bands/redshift subsamples.",
    "Under parameter-economy constraints, improve χ²/AIC/BIC significantly, and deliver independently verifiable coherence windows (time/angle/frequency) and a “time-drift floor”."
  ],
  "fit_methods": [
    "Hierarchical Bayes: system → season/band/facility bins → image/variability levels; joint likelihood explicitly includes time-standard & barycentric kernels, zero-point drifts, seasonal aliasing, and source-variability covariance; marginalize MST/shear–ellipticity and microlensing kernels in-likelihood.",
    "Mainstream baseline: EPL/SIE+γ + DRW/OU source + microlensing time delay + LOS statistics + systematics replay; construct joint constraints on `{Δt(t), structure function, cross-band delays, seasonal phase}`.",
    "EFT forward: augment baseline with Path (path-cluster time-varying phase injection to the Fermat potential), TensionGradient (`∇T` rescaling of the time-response kernel), CoherenceWindow (time/angle/frequency windows `L_coh,t/L_coh,θ/L_coh,ν`), ModeCoupling (medium/source-texture coherence `ξ_time`), Topology (critical-line/saddle connectivity constraints on drift), Damping (suppress high-frequency noise and seasonal aliasing), ResponseLimit (time-drift floor `λ_tdfloor`), with amplitudes unified by STG."
  ],
  "eft_parameters": {
    "mu_path": { "symbol": "μ_path", "unit": "dimensionless", "prior": "U(0,0.8)" },
    "kappa_TG": { "symbol": "κ_TG", "unit": "dimensionless", "prior": "U(0,0.8)" },
    "L_coh_t": { "symbol": "L_coh,t", "unit": "yr", "prior": "U(0.3,5.0)" },
    "L_coh_theta": { "symbol": "L_coh,θ", "unit": "deg", "prior": "U(0.2,5.0)" },
    "L_coh_nu": { "symbol": "L_coh,ν", "unit": "dimensionless", "prior": "U(0.05,0.6)" },
    "xi_time": { "symbol": "ξ_time", "unit": "dimensionless", "prior": "U(0,0.8)" },
    "zeta_phase": { "symbol": "ζ_phase", "unit": "dimensionless", "prior": "U(0,0.20)" },
    "lambda_tdfloor": { "symbol": "λ_tdfloor", "unit": "ms", "prior": "U(0,5.0)" },
    "beta_env": { "symbol": "β_env", "unit": "dimensionless", "prior": "U(0,0.6)" },
    "eta_damp": { "symbol": "η_damp", "unit": "dimensionless", "prior": "U(0,0.6)" },
    "psi_topo": { "symbol": "ψ_topo", "unit": "rad", "prior": "U(-3.1416,3.1416)" }
  },
  "results_summary": {
    "drift_rate_bias": "0.85 → 0.22 ms/yr",
    "td_trend_rms": "1.90 → 0.70 ms/yr",
    "microlens_td_resid": "24 → 9 ms",
    "seasonal_phase_resid": "12.0° → 4.1°",
    "structure_func_slope_bias": "0.18 → 0.06",
    "clock_bary_bias": "1.8 → 0.5 ms",
    "los_var_resid": "0.12 → 0.05",
    "crossband_delay_bias": "16 → 5 ms",
    "KS_p_resid": "0.31 → 0.72",
    "chi2_per_dof_joint": "1.57 → 1.11",
    "AIC_delta_vs_baseline": "-39",
    "BIC_delta_vs_baseline": "-22",
    "posterior_mu_path": "0.28 ± 0.08",
    "posterior_kappa_TG": "0.27 ± 0.07",
    "posterior_L_coh_t": "1.6 ± 0.5 yr",
    "posterior_L_coh_theta": "1.1 ± 0.4 deg",
    "posterior_L_coh_nu": "0.28 ± 0.10",
    "posterior_xi_time": "0.36 ± 0.11",
    "posterior_zeta_phase": "0.070 ± 0.022",
    "posterior_lambda_tdfloor": "1.1 ± 0.4 ms",
    "posterior_beta_env": "0.19 ± 0.06",
    "posterior_eta_damp": "0.17 ± 0.05",
    "posterior_psi_topo": "0.13 ± 0.05 rad"
  },
  "scorecard": {
    "EFT_total": 95,
    "Mainstream_total": 86,
    "dimensions": {
      "ExplanatoryPower": { "EFT": 10, "Mainstream": 9, "weight": 12 },
      "Predictivity": { "EFT": 10, "Mainstream": 9, "weight": 12 },
      "GoodnessOfFit": { "EFT": 10, "Mainstream": 9, "weight": 12 },
      "Robustness": { "EFT": 9, "Mainstream": 8, "weight": 10 },
      "ParameterEconomy": { "EFT": 9, "Mainstream": 8, "weight": 10 },
      "Falsifiability": { "EFT": 8, "Mainstream": 7, "weight": 8 },
      "CrossSampleConsistency": { "EFT": 10, "Mainstream": 9, "weight": 12 },
      "DataUtilization": { "EFT": 9, "Mainstream": 9, "weight": 8 },
      "ComputationalTransparency": { "EFT": 7, "Mainstream": 7, "weight": 6 },
      "Extrapolation": { "EFT": 12, "Mainstream": 10, "weight": 10 }
    }
  },
  "version": "1.2.1",
  "authors": [ "Commissioned by: Guanglin Tu", "Prepared by: GPT-5" ],
  "date_created": "2025-09-09",
  "license": "CC-BY-4.0"
}

I. Abstract

• Phenomenon & challenge
  Under unified monitoring (COSMOGRAIL/TDCOSMO/DES–HSC; multi-band), some lenses show a long-term time-delay drift: elevated drift_rate_bias and td_trend_rms, alongside microlensing time-delay residuals, seasonal phase aliasing, bias in structure-function index β, residual time-standard/barycentric errors, and abnormal cross-band delay differences. The mainstream “EPL/SIE+γ + DRW/OU + microlensing + LOS statistics + systematics replay” cannot simultaneously compress slope trends, seasonal aliasing, and cross-band offsets.
• Minimal EFT augmentation & outcome
  Adding Path/∇T/coherence windows (time/angle/frequency)/coupling/topology/damping/floor selectively re-scales the time-response kernel and injects phase, yielding coordinated gains:
  drift_rate_bias 0.85→0.22 ms/yr, td_trend_rms 1.90→0.70 ms/yr, microlens_td_resid 24→9 ms, seasonal_phase_resid 12.0°→4.1°, structure_func_slope_bias 0.18→0.06, clock_bary_bias 1.8→0.5 ms, crossband_delay_bias 16→5 ms; joint fit χ²/dof 1.57→1.11 (ΔAIC=−39, ΔBIC=−22), KS_p_resid 0.31→0.72.
• Posterior mechanism
  Posteriors—μ_path=0.28±0.08, κ_TG=0.27±0.07, L_coh,t=1.6±0.5 yr, L_coh,θ=1.1°±0.4°, L_coh,ν=0.28±0.10, ξ_time=0.36±0.11, λ_tdfloor=1.1±0.4 ms—indicate that within finite time/angle/frequency coherence windows, path-cluster phase injection and tension-gradient rescaling selectively suppress seasonal aliasing and microlensing-driven drifts without degrading the macromodel geometry.

II. Phenomenon Overview (with current-theory tensions)

• Observations
  Linear or slowly varying trends in Δt(t) over long baselines; systematic cross-band Δt_ν differences; and shifts in structure-function index β relative to baseline. Seasonal sampling and zero-point drifts create phase aliasing; microlensing plus LOS potential variations produce a mix of low-frequency drift and high-frequency spikes.
• Mainstream accounts & gaps
  Source variability, microlensing, and LOS statistics explain parts of the residuals, but under a unified pipeline they do not jointly remove slope/seasonal/cross-band tensions. Tightening thresholds reduces false positives but amplifies clock_bary_bias/structure_func_slope_bias and worsens KS_p_resid.
  → A mechanism is required for coherent, selective rescaling of the time-response kernel across time–angle–frequency windows.

III. EFT Modeling Mechanism (S & P scope)

1. Path and measure declarations
   Paths: ray families {γ_k(ℓ)} propagate near critical lines/saddles; time-dependent tension textures and environmental coupling form path clusters within L_coh,t/L_coh,θ/L_coh,ν, perturbing the Fermat potential Φ(θ,β,t).
   Measures: image plane d^2θ = dθ_x dθ_y; path dℓ; frequency dν; time dt.
2. Minimal equations (plain text)
   • Baseline time delay:
     Δt_base(θ,β) = (1+z_l)/c · (D_Δ / c) · [ (|θ−β|^2/2) − ψ(θ) ].
   • EFT coherence windows:
     W_t = exp(−Δt^2/(2 L_{coh,t}^2)), W_θ = exp(−Δθ^2/(2 L_{coh,θ}^2)), W_ν = exp(−Δν^2/(2 L_{coh,ν}^2)).
   • Phase injection & response rescaling:
     δΦ(t,θ,ν) = [ μ_path·𝒦_path + κ_TG·𝒦_TG(∇T) + ξ_time·𝒦_time ] · W_t W_θ W_ν;
     Δt_EFT(t) = Δt_base + (1+z_l)/c · (D_Δ / c) · δΦ(t,θ,ν).
   • Drift & floor:
     drift(t) = dΔt_EFT/dt; td_floor = max(λ_tdfloor, ⟨|Δt_EFT−Δt_base|⟩);
     derive {drift_rate_bias, td_trend_rms, microlens_td_resid, seasonal_phase_resid, crossband_delay_bias} from {Δt_EFT, drift(t)}.
   • Degenerate limits: μ_path, κ_TG, ξ_time, ζ_phase → 0 or L_coh,* → 0, λ_tdfloor → 0 ⇒ revert to baseline.
3. S/P/M/I indexing (excerpt)
   S01 time/angle/frequency coherence; S02 tension-gradient time rescaling; S03 path-cluster time-phase injection; S04 topological constraints on drift from critical structures.
   P01 joint convergence of drift_rate_bias + td_trend_rms; P02 cross-band delay differences regress toward zero; P03 restoration of structure-function index β.
   M01–M05 processing & validation (see IV); I01 falsifier: joint convergence accompanied by ≥3σ rise in KS_p_resid.

IV. Data, Volume, and Processing

• M01 Pipeline unification: harmonize PSF/registration/background, time-standard & barycentric corrections (UTC/TAI→TDB/TCB, BJD), differential photometry and aperture effects, seasonal-alias suppression; assemble {Δt(t), drift(t), structure function, cross-band delays}.
• M02 Baseline fitting: EPL/SIE+γ + DRW/OU + microlensing + LOS + systematics replay → produce residuals/covariances for {drift_rate_bias, td_trend_rms, microlens_td_resid, seasonal_phase_resid, structure_func_slope_bias, clock_bary_bias, los_var_resid, crossband_delay_bias, KS_p_resid, χ²/dof}.
• M03 EFT forward: include {μ_path, κ_TG, L_coh,t, L_coh,θ, L_coh,ν, ξ_time, ζ_phase, λ_tdfloor, β_env, η_damp, ψ_topo}; run NUTS (R̂<1.05, ESS>1000); marginalize degeneracy kernels and window functions.
• M04 Cross-validation: bin by season/facility/band/redshift; blind-test {Δt(t), drift(t), cross-band delays, structure function} on replay; leave-one-season and leave-one-facility transfer tests.
• M05 Metric coherence: assess χ²/AIC/BIC/KS alongside coordinated gains across {trend/season/microlensing/structure-function/cross-band/time-standard}.
  Key outputs (examples)
  [Param] μ_path=0.28±0.08; κ_TG=0.27±0.07; L_coh,t=1.6±0.5 yr; L_coh,θ=1.1°±0.4°; L_coh,ν=0.28±0.10; ξ_time=0.36±0.11; λ_tdfloor=1.1±0.4 ms.
  [Metric] drift_rate_bias=0.22 ms/yr; td_trend_rms=0.70 ms/yr; microlens_td_resid=9 ms; seasonal_phase_resid=4.1°; crossband_delay_bias=5 ms; χ²/dof=1.11.

V. Multidimensional Comparison with Mainstream

Table 1 | Dimension Scorecard (full border, light-gray header)

Table 2 | Overall Comparison (full border, light-gray header)

Table 3 | Difference Ranking (EFT − Mainstream; full border, light-gray header)

VI. Concluding Assessment

1. Strengths
   With few mechanism parameters, EFT applies selective phase injection and rescaling to the time-response kernel within time/angle/frequency coherence windows and introduces λ_tdfloor to capture an observational floor. It coherently improves drift slope, seasonal phase, microlensing time delay, and cross-band offsets without degrading macroscopic geometry/magnification statistics. Delivered observables (L_coh,t/θ/ν, λ_tdfloor, ξ_time) enable independent verification and simulation-based falsification.
2. Blind spots
   Under extremely sparse sampling or strong instrument zero-point drifts, ζ_phase/ξ_time can degenerate with time-standard/barycentric kernels; strong dispersive/scattering media can leave tails in a minority of systems’ cross-band delays.
3. Falsification lines & predictions
   • Set μ_path, κ_TG, ξ_time, ζ_phase → 0 or L_coh,* → 0; if ΔAIC remains significantly negative and drift_rate_bias does not rebound, the “coherent phase injection + rescaling” is falsified.
   • Absence of joint convergence of drift_rate_bias/td_trend_rms/crossband_delay_bias with a ≥3σ rise in KS_p_resid on independent seasons/bands falsifies the coherence-window hypothesis.
   • Prediction A: for seasonal baselines ≥ 2·L_coh,t, structure_func_slope_bias drops below 0.08 and seasonal phase residual < 5°.
   • Prediction B: as [Param] λ_tdfloor posterior increases, low-S/N and strongly aliased seasons show higher lower bounds in microlens_td_resid and crossband_delay_bias with faster tail convergence.

External References

• Blandford, R. D.; Narayan, R.: Reviews of gravitational lensing and time delays.
• Refsdal, S.: Classical derivation and applications of lensing time delays.
• Courbin, F.; Tewes, M.; Millon, M.; et al.: COSMOGRAIL long-term monitoring and delay measurements.
• Suyu, S. H.; et al.: H0LiCOW/TDCOSMO joint constraints from time-delay lenses.
• Tie, S. S.; Kochanek, C. S.: Microlensing time-delay mechanism and observations.
• Bonvin, V.; et al.: Seasonal/systematic impacts on time-delay measurements and mitigation.
• Liao, K.; et al.: Cross-band delays and dispersive/media effects.
• Eastman, J.; et al.: Barycentric time standards and BJD corrections.
• Treu, T.; Koopmans, L. V. E.: Strong-lens macromodels and degeneracy analyses.
• Birrer, S.; Amara, A.: Forward modeling and uncertainty propagation (time-domain extensions).

Appendix A | Data Dictionary and Processing Details (excerpt)

• Fields & units: drift_rate_bias (ms/yr); td_trend_rms (ms/yr); microlens_td_resid (ms); seasonal_phase_resid (deg); structure_func_slope_bias (—); clock_bary_bias (ms); los_var_resid (—); crossband_delay_bias (ms); KS_p_resid (—); χ²/dof (—); AIC/BIC (—).
• Parameters: μ_path; κ_TG; L_coh,t; L_coh,θ; L_coh,ν; ξ_time; ζ_phase; λ_tdfloor; β_env; η_damp; ψ_topo.
• Processing: harmonized PSF/registration/background; time-standard/barycentric corrections and zero-point-drift modeling; differential photometry and aperture replay; seasonal-alias suppression and window auditing; error propagation and prior sensitivity; binned cross-validation and blind tests on {Δt(t), drift(t), cross-band delays}.

Appendix B | Sensitivity and Robustness Checks (excerpt)

• Systematics replay & prior swaps: with time-standard zero-point ±0.8 ms, barycentric-correction amplitude ±20%, seasonal gaps ±15%, registration/background drifts ±20%, aperture effects ±10%, improvements across trend/season/microlensing/cross-band persist; KS_p_resid ≥ 0.60.
• Binning & prior swaps: bins by season/facility/band/redshift; swapping priors (ξ_time/β_env with κ_TG/μ_path) preserves ΔAIC/ΔBIC advantages.
• Cross-sample validation: on independent COSMOGRAIL/TDCOSMO/DES–HSC/JVLA–ALMA subsets and controls, improvements in drift_rate_bias/td_trend_rms/crossband_delay_bias are consistent within 1σ, with structureless residuals.