V2.105: The 4D Trace Anomaly Identity in de Sitter

Status: COMPLETE — NEGATIVE RESULT (but with important positive insights)

Headline

The two routes to Lambda are NOT equivalent — and this is a feature, not a bug. Semiclassical Einstein equations sourced by the trace anomaly stress tensor give an O(H⁴) correction with the wrong sign for a positive cosmological constant. The Cai-Kim thermodynamic route gives an O(H²) term — Lambda itself — with the correct sign. The identity δ = −a (Casini-Huerta 2012) ensures both routes use the same anomaly coefficient, but they process it at fundamentally different orders.

The Question

An external reviewer suggested proving a 4D identity: that the gravitational field equations sourced by ⟨T_ab⟩_anomaly in de Sitter give the same equation as the Cai-Kim first law with log-corrected entropy — not just the same scaling, but the same coefficient. This would be the 4D version of the exact 1+1D Casimir identity.

The Answer

They do NOT give the same equation. The comparison reveals:

Property	Route A (Semiclassical)	Route B (Thermodynamic)
Starting point	⟨T_ab⟩ = ρ_anom g_ab	S = αA + δ ln(A)
Key equation	G_ab + Λ_bare g_ab = 8πG ⟨T_ab⟩	−dE = T dS at horizon
Order	O(H⁴/M_Pl²) correction	O(H²) = Λ itself
Sign	ρ_anom = −6aH⁴ < 0 (wrong for Λ > 0)	Λ = |δ|H²/(2α) > 0 (correct)
de Sitter with Λ_bare = 0?	NO (no solution exists)	YES (self-consistency: a = 6α)

Key Findings

Finding 1: The sign problem kills Route A

The trace anomaly in de Sitter:

• ⟨T^a_a⟩ = a × E₄ = 24aH⁴ (with W² = 0, E₄ = 24H⁴)
• By maximal symmetry: ⟨T_ab⟩ = ρ_anom g_ab with ρ_anom = −6aH⁴

The semiclassical Friedmann equation:

• 3H² = Λ_bare + (2π/α)(−6aH⁴) = Λ_bare − 12πaH⁴/α

Setting Λ_bare = 0: 3H² = −12πaH⁴/α → H² < 0. No de Sitter solution.

Finding 2: O(H⁴) vs O(H²) — different physics

Route A contributes at O(H⁴/M_Pl²) — this is Starobinsky’s (1980) anomaly-driven inflation correction. It’s negligible at late cosmological times.

Route B contributes at O(H²) — this IS Lambda. It’s the dominant term.

Numerically: |Λ_anomaly^(A)| / Λ^(B) = 8πR_SM ≈ 16.5 at SM self-consistency. They operate at completely different scales.

Finding 3: The actual 4D identity is δ = −a (already proven)

The correct 4D generalization of the 1+1D Casimir identity is:

δ = −a (Casini-Huerta 2012)

This says: the UV-finite log coefficient of entanglement entropy for a sphere equals minus the type-A trace anomaly coefficient. This is PROVEN via the connection to the F-theorem and strong subadditivity.

The 1+1D identity says: the subleading entropy correction IS the Casimir energy. The 3+1D identity says: the log entropy coefficient IS the trace anomaly coefficient.

Both connect entropy to the anomaly. The difference is in how this connection enters the field equations.

Finding 4: The real remaining gap is the Cai-Kim first law

What IS proven:

• ✓ δ = −a for spherical entangling surfaces (Casini-Huerta 2012)
• ✓ a-theorem: ‘a’ decreases under RG flow (Komargodski-Schwimmer 2011)
• ✓ Jacobson: G = 1/(4α) from area law (1995)
• ✓ V2.67: δ = −1/90 confirmed to 1% on lattice

What is NOT proven:

• ✗ Λ_bare = 0 (assumption)
• ✗ Cai-Kim first law at cosmological horizon (physical framework, not theorem)
• ✗ The O(H²) thermodynamic Λ is the only Λ

Finding 5: The Euclidean path is the closest to a proof

The anomaly effective action W[g] on the Euclidean S⁴ generates:

1. The stress tensor ⟨T_ab⟩ = (2/√g) δW/δg^ab → O(H⁴) correction
2. The entropy S = −∂W/∂T via the partition function → S = αA − a ln(A)

Both come from the SAME functional W[g]. The entropy route (2) gives Lambda through the first law; the stress tensor route (1) gives the wrong sign. The entropy is more fundamental.

What This Means for the Paper

The paper should not attempt to claim that Routes A and B are equivalent. They aren’t. Instead, the paper should:

1. Emphasize that the entropy route is more fundamental (Jacobson’s original insight)
2. Cite Casini-Huerta δ = −a as the 4D identity (it IS the identity)
3. Note that the O(H⁴) semiclassical correction exists but is irrelevant at late times
4. Acknowledge that the Cai-Kim first law is the irreducible physical assumption

What Would Actually Be a Breakthrough

The true breakthrough would be proving that the Cai-Kim first law is exact at the cosmological horizon — i.e., that −dE = TdS holds as a fundamental identity (not an approximation) with the full log-corrected entropy. This is equivalent to proving Jacobson’s thermodynamic gravity conjecture, which is one of the deepest open problems in theoretical physics.

A more tractable step: proving that the Euclidean anomaly effective action on the conical S⁴ (the replica trick geometry for the cosmological horizon) generates S = αA − a ln(A) AND that the resulting Wald-like first law gives Lambda = a H²/(2α). This would be a genuine contribution and connects to the Fursaev (1995) / Sen (2013) program of computing log corrections from the Euclidean effective action.

Numerical Summary

Quantity	Value
E₄ (de Sitter)	24H⁴
W² (de Sitter)	0
ρ_anom (Route A)	−6aH⁴
Λ (Route B)	|δ|H²/(2α)
Self-consistency factor	f = 6
R_SM = |δ_SM|/(6α_SM)	0.657
Λ_SM/Λ_obs	0.960 (4% low)
Λ_SM+grav/Λ_obs	1.060 (6% high)
Midpoint	1.010 (1% off)
Route A magnitude / Route B	~16.5 (confirms different orders)

Files

• src/conventions.py — Sign convention tracking and curvature invariants
• src/anomaly_stress_tensor.py — Route A: trace anomaly → Friedmann
• src/cai_kim_derivation.py — Route B: log entropy → Cai-Kim → Friedmann
• src/identity_comparison.py — Route C: comparison and identity analysis
• tests/test_identity.py — 11 tests, all passing
• results/identity_verification.json — Full results