V2.138 - DESI Dark Energy Equation of State — Falsification Test
V2.138: DESI Dark Energy Equation of State — Falsification Test
Status: COMPLETE
Question
The entanglement framework makes a hard prediction: the dark energy equation of state is w = -1 exactly at all redshifts. The cosmological constant is a static vacuum energy derived from the trace anomaly log correction to entanglement entropy — there is no dynamical dark energy. DESI BAO data (DR1 2024, DR2 2025) shows a growing preference for w₀waCDM over ΛCDM. How serious is this tension?
Method
- Compile published DESI constraints on w₀ and wₐ from the CPL parameterization w(z) = w₀ + wₐ z/(1+z)
- Compute Mahalanobis distance (tension in σ) from ΛCDM point (w₀=-1, wₐ=0) using full 2D covariance
- Bayesian evidence via Savage-Dickey density ratio
- w(z) evolution trajectories and luminosity distance differences
- Project tensions to future experiments (DESI DR3, Euclid, LSST)
Results
Tension with ΛCDM
| Dataset | Tension (σ) | p-value | Published Δχ² |
|---|---|---|---|
| DESI DR1 + CMB + PantheonPlus | 2.0σ | 0.049 | 9.0 |
| DESI DR1 + CMB + Union3 | 3.7σ | 2.1 × 10⁻⁴ | 7.2 |
| DESI DR2 + CMB + PantheonPlus | 4.2σ | 3.2 × 10⁻⁵ | 12.4 |
| DESI DR2 + CMB + DESY5 | 3.3σ | 9.5 × 10⁻⁴ | 9.8 |
The tension has grown from ~2σ (DR1 + PantheonPlus) to 3.3–4.2σ (DR2), depending on the supernova dataset. The spread between PantheonPlus (4.2σ) and DESY5 (3.3σ) indicates residual systematic uncertainties in supernova calibration.
Bayesian Evidence
Using the Savage-Dickey density ratio with flat priors w₀ ∈ [-2, 0], wₐ ∈ [-3, 3]:
| Dataset | B(ΛCDM/w₀wₐ) | log₁₀ B | Interpretation |
|---|---|---|---|
| DESI DR1 + CMB + PantheonPlus | 1.114 | +0.05 | Weak evidence for ΛCDM |
| DESI DR1 + CMB + Union3 | 0.041 | -1.39 | Strong evidence for w₀wₐCDM |
| DESI DR2 + CMB + PantheonPlus | 0.009 | -2.03 | Decisive evidence for w₀wₐCDM |
| DESI DR2 + CMB + DESY5 | 0.301 | -0.52 | Moderate evidence for w₀wₐCDM |
The Bayesian picture is mixed. DR2 + PantheonPlus decisively prefers dynamical dark energy (B = 0.009), but DR2 + DESY5 gives only moderate preference (B = 0.3). This SNe-dependence is a red flag for systematics.
Phantom Divide Crossing
All DESI best fits cross the phantom divide (w = -1) at z ≈ 0.5–0.6:
- w(z=0) ≈ -0.75 (phantom, w > -1)
- w(z→∞) ≈ -1.43 (quintessence-like, w < -1)
This phantom crossing is theoretically problematic — in standard scalar field theories, crossing w = -1 requires either a ghost or multiple fields. This may indicate the signal is a systematic artifact rather than real physics.
Luminosity Distance Differences
The DESI best-fit w₀wₐCDM models predict distances that differ from ΛCDM by:
| z | DR2+Pantheon | DR2+DESY5 |
|---|---|---|
| 0.5 | -2.6% | -2.0% |
| 1.0 | -2.5% | -1.9% |
| 2.0 | -1.7% | -1.4% |
These are small (1–3%) deviations — within the range where supernova calibration systematics matter.
Falsification Projections
If the current DESI DR2 + CMB + PantheonPlus central values hold:
| Experiment | σ(w₀) | σ(wₐ) | Projected tension |
|---|---|---|---|
| Current (DR2) | 0.055 | 0.18 | 4.2σ |
| DESI DR3 (2026–27) | ~0.038 | ~0.13 | ~6.2σ |
| Euclid DR1 (2027) | ~0.03 | ~0.10 | ~8σ |
| LSST Year 3 (2028) | ~0.025 | ~0.08 | >10σ |
If the central values don’t move, DESI DR3 alone would achieve 5σ falsification within 1–2 years. Euclid and LSST would make it definitive.
However — if the apparent tension is driven by supernova systematics (suggested by the PantheonPlus vs DESY5 discrepancy), the central values WILL move toward w = -1 as systematics are resolved, and the tension will stabilize or decrease.
Honest Assessment
What this means for the framework
The DESI data is the most serious challenge the entanglement framework faces. At 3.3–4.2σ, the tension with w = -1 is substantial but not definitive. The situation is genuinely uncertain:
Reasons for concern:
- The tension is growing with more data (DR1 → DR2)
- DR2 + PantheonPlus gives 4.2σ and decisive Bayesian evidence against ΛCDM
- If real, this would definitively falsify the framework — there is no escape route
Reasons for cautious optimism:
- The PantheonPlus vs DESY5 spread (4.2σ vs 3.3σ) indicates unresolved SNe systematics
- The phantom crossing at z ≈ 0.5 is theoretically problematic — few physical models predict this
- BAO data alone (without SNe) gives weaker tension
- The “trend” from DR1 → DR2 is partially driven by better statistics on the SAME underlying data, not independent confirmation
- Previous “discoveries” of dynamical dark energy have evaporated when systematics were resolved (DES Y1, Planck 2015)
The bottom line
The framework is NOT falsified but is under genuine pressure. The prediction w = -1 is falsifiable in principle and the data is pushing against it. DESI DR3 (expected 2026–27) will be decisive:
- If central values hold and errors shrink → 5σ+ falsification
- If central values drift toward w = -1 → the framework survives
This is exactly what a good scientific prediction looks like: it makes a sharp claim that upcoming experiments can definitively test.
What the framework predicts will happen
If the entanglement derivation is correct, then w = -1 exactly. The current DESI tension must be explained by:
- Supernova calibration systematics (most likely — see PantheonPlus vs DESY5 discrepancy)
- BAO modeling assumptions (possible — see “full shape” vs “compressed” BAO analyses)
- CMB lensing anomaly propagation into w₀-wₐ (known issue)
The framework predicts that as these systematics are resolved, the tension will decrease. This is a testable meta-prediction.
Files
run_experiment.py: Main experiment driver (7 phases)src/desi_constraints.py: Published DESI DR1/DR2 constraints, Bayesian evidence, projectionssrc/w_evolution.py: w(z) trajectories, distance computations, phantom crossing diagnosticstests/test_desi.py: 13 tests (all pass)results/results.json: Full numerical data