E3d-v2 — powered pair-bridge measurement
CMASS-N z 0.43–0.7 (876639 pairs, transverse 6.0–14.0 Mpc/h, LOS < 6.0). Estimator: 2.0′-disc-averaged y at the midpoint minus the mean over ring controls at ±60/±90/±120° around each galaxy. Bootstrap errors over pairs.
| map | pairs used | bridge excess (y) | error | SNR |
|---|---|---|---|---|
| ACT | 361057 | 2.278e-08 | 4.348e-09 | 5.24 |
| PLANCK | 876639 | 2.966e-08 | 1.536e-09 | 19.31 |
Verdict — gate passed
- Model B established at ACT depth: 5.24σ (gate: ≥3σ). At 1.6′ resolution second-halo beam leakage is negligible, so the on-axis excess is inter-halo gas: the filament bridge. Amplitude 2.28×10⁻⁸, consistent with Planck’s 2.97×10⁻⁸ (19σ, small residual leakage expected at 10′) and with published LRG-pair bridge measurements (~1–2×10⁻⁸ for narrower separation windows).
- Two instruments, one geometric estimator, agreeing amplitudes — the between-halos question that E3b/E3c could only bound is now a detection under the design matched to the instrument.
- Remaining caveats for a publication-grade number: model-based two-halo subtraction (our ring estimator is geometric), dust/CIB null tests (145 vs 90 GHz, NILC vs MILCA), and bootstrap over sky patches rather than pairs (pairs overlap on the sky).