IR-fixed Euclidean vacuum for linearized gravity on de Sitter space

This paper constructs a gauge-invariant and Hadamard Euclidean vacuum for linearized gravity on global de Sitter space using Calderón projectors, demonstrating that while the initial state is not positive-definite on the entire phase space, a low-energy modification yields a well-defined Hadamard state.

The Gist

The Big Picture: Trying to Tune a Cosmic Radio

Imagine the universe is a giant, expanding balloon (this is De Sitter space). Physicists want to understand the tiny ripples on this balloon—these are gravitational waves or "linearized gravity." To do this, they need to set up a "quantum radio station" that broadcasts the rules of how these ripples behave.

The goal of this paper is to find the perfect "station" (called a vacuum state) that is:

1. Stable: It doesn't explode or give negative energy (Positivity).
2. Symmetric: It looks the same from all angles and respects the laws of physics (Gauge Invariance).
3. Realistic: It behaves correctly at both very small scales (UV) and very large scales (IR).

The Problem: The "Ghost" Signal

The authors tried to build this station using a clever trick called Wick Rotation.

• The Analogy: Imagine you want to study a stormy ocean (Lorentzian space, where time flows forward). It's chaotic and hard to calculate. So, you imagine the ocean is frozen into a solid, smooth sphere of ice (Euclidean space). It's much easier to solve math problems on the ice. Once you solve it, you "melt" the ice back into water to get your answer.

The authors took the solution from the "ice sphere" (the 4-sphere) and melted it back into the "ocean" (De Sitter space). This is the standard Euclidean Vacuum (often called the Bunch-Davies vacuum in physics).

However, they found a glitch.
When they melted the ice back, the resulting radio station had a static noise that made the signal impossible to listen to. Specifically:

• The Positivity Issue: In quantum physics, energy must be positive. If you have "negative energy," the universe becomes unstable. The standard Euclidean vacuum produced a tiny, finite number of "ghost modes" (mathematical errors) that had negative energy.
• The Gauge Issue: Gravity has a lot of "redundant" ways to describe the same thing (like describing a car's position by its GPS coordinates vs. its street address). The math needs to ignore these redundancies. The standard vacuum failed to ignore a specific set of these redundancies properly.

It was like tuning a radio to a station, but the signal was so corrupted by static (negative energy) and ghost voices (gauge errors) that you couldn't hear the music.

The Solution: The "Filter" (IR-Fix)

The authors didn't throw the whole idea away. Instead, they realized the problem wasn't the whole station, just a few specific "bad frequencies."

1. Identifying the Culprits: They discovered that the negative energy and gauge errors were coming from a very small, finite group of mathematical modes (specifically related to "Killing 1-forms," which are like special, rigid patterns the universe can vibrate in).
2. The Fix: They created a mathematical filter (a projection operator).
   • The Analogy: Imagine you have a bucket of water with a few rotten apples floating in it. Instead of throwing away the whole bucket of water (which would mean giving up on the theory), you just use a strainer to scoop out the rotten apples.
   • They scooped out those specific "bad modes" (the $E_{TT,4}$ subspace) and kept everything else.

The Result: A New, Clean Vacuum

By removing just those few bad apples, they created a Modified Euclidean Vacuum.

• It works: The new signal is clean. All energy is positive.
• It's symmetric: It respects the rules of the universe, though it loses a tiny bit of perfect symmetry (it's no longer invariant under every possible movement of the universe, only the ones that keep a specific "slice" of time fixed).
• It's rigorous: Unlike previous attempts that tried to fix this by hand-waving or infinite adjustments, this paper provides a strict, mathematical proof that the fix works.

Why This Matters

Before this paper, there was a nagging doubt in the physics community: Does the standard "Bunch-Davies" vacuum actually work for gravity on an expanding universe?

This paper says: "Not exactly, but we can fix it easily."

They showed that the standard method produces a "pseudo-state" (a broken signal) and provided the precise mathematical tool to repair it. This is a crucial step for anyone trying to do serious calculations about the early universe, black holes, or the quantum nature of gravity, ensuring that the math doesn't collapse under its own contradictions.

Summary in One Sentence

The authors found that the standard mathematical recipe for the "quantum vacuum" of gravity on an expanding universe produces a few "negative energy" glitches, but they successfully built a filter to remove those glitches, resulting in a stable, physically valid theory.

Technical Summary

1. Problem Statement

The paper addresses the mathematical challenges in quantizing linearized gravity on global de Sitter space ($dS_4$). Specifically, it investigates the validity of the Euclidean (or Bunch–Davies) vacuum, a state constructed by Wick-rotating the Lorentzian theory to a Euclidean theory on the 4-sphere ($S_4$), obtaining the Green's function, and rotating back.

While this construction is standard for scalar fields and has been rigorously established for other gauge theories (like Maxwell or Yang-Mills) on compact Cauchy surfaces, it faces specific obstructions in linearized gravity:

1. Positivity Failure: The resulting two-point function (covariance) fails to be positive definite on the entire physical phase space.
2. Gauge Invariance Issues: The state may not satisfy the required "strong" gauge invariance conditions necessary for a well-defined quantum state on the quotient space of physical solutions.
3. IR Divergences: Previous literature suggested these issues were related to infrared (IR) divergences in mode expansions, but the authors aim to provide a rigorous, finite-dimensional resolution.

2. Methodology

The authors employ a rigorous microlocal analysis approach, avoiding the ambiguities of formal mode expansions. Their methodology consists of the following key steps:

• Wick Rotation and Geometry: They treat de Sitter space ($dS_4 = \mathbb{R} \times S^3$) and the Euclidean sphere ($S^4$) as analytic continuations of each other. The Lorentzian metric $g = -dt^2 + \cosh^2(t)h$ is rotated to the Riemannian metric $\tilde{g} = ds^2 + \cos^2(s)h$.
• Calderón Projectors: Instead of relying on mode expansions (which may not converge rigorously in this context), they define the vacuum state via Calderón projectors. These are projections onto the subspace of Cauchy data of $L^2$ solutions to the elliptic operator $\tilde{D}_2$ (the Wick-rotated linearized Einstein operator) on the hemispheres $\Omega_\pm = \{\pm s > 0\} \subset S^4$.
• TT-Gauge Formalism: To handle the gauge freedom, they work in the Transverse-Traceless (TT) gauge. They define the physical phase space as a quotient of Cauchy data spaces:
  • $E_{TT}$: Space of Cauchy data for traceless, harmonic solutions.
  • $F_{TT}$: Subspace representing residual gauge freedom.
  • Physical Space: $E_{TT} / F_{TT}$.
• Spectral Analysis: They perform a detailed spectral analysis of the Wick-rotated operators on $S^4$, specifically utilizing the decomposition of tensor fields into eigenmodes of the Lichnerowicz Laplacians. This allows them to isolate specific subspaces where the positivity condition fails.

3. Key Contributions and Results

A. Identification of the Obstruction (Theorem 1.2)

The authors prove that the standard Euclidean Green's function on $S^4$, when Wick-rotated back, does not define a valid quantum state for linearized gravity on $dS_4$.

• Hadamard Condition: The state satisfies the Hadamard condition (correct UV behavior).
• Weak Gauge Invariance: It satisfies weak gauge invariance.
• Positivity Failure: It fails the positivity condition on the physical phase space. Specifically, the covariance is negative on a specific 6-dimensional subspace ($E_{TT,4}$) of the phase space.
• Strong Gauge Invariance Failure: The state is not "strongly" gauge invariant; the covariance does not vanish on the entire gauge orbit, only on a specific subspace related to Killing 1-forms.

The subspace $E_{TT,4}$ is explicitly identified as the space of Cauchy data corresponding to Killing 1-forms on $S^4$ (which are related to linearization instabilities). This is a finite-dimensional obstruction, distinct from the infinite-dimensional IR divergences often discussed in physics literature.

B. Construction of the Modified Vacuum (Theorem 1.3)

The authors propose a mathematically rigorous finite-dimensional modification to resolve the positivity issue:

• They define a projection operator $\pi$ that projects the phase space $E_{TT}$ onto the subspace $E_{TT, \text{gauge}} \oplus F_{TT, \text{gauge}}$, effectively removing the problematic 6-dimensional subspace $E_{TT,4}$.
• The Modified Euclidean Vacuum ($\omega_{\text{mod}}$) is defined by composing the original pseudo-state with this projection.
• Properties of $\omega_{\text{mod}}$:
  1. It is a well-defined Hadamard state.
  2. It is positive on the entire physical phase space.
  3. It is fully gauge invariant (satisfies strong gauge invariance).
  4. It is invariant under the subgroup $O(4)$ (isometries preserving the $t=0$ slice).
  5. Symmetry Breaking: It is not invariant under the full de Sitter isometry group $SO^\uparrow(1,4)$. The modification breaks the full symmetry because the projection $\pi$ does not commute with the generators of boosts (which mix the problematic modes with the physical ones).

C. Analogy with Maxwell Fields

In Appendix B, the authors demonstrate that this phenomenon is not unique to gravity. They perform a similar analysis for Maxwell fields on de Sitter space, finding an analogous finite-dimensional obstruction ($E_0$) related to constant modes, which is resolved by a similar projection.

4. Significance

• Rigorous Resolution of IR Issues: The paper clarifies that the "IR divergence" in the Euclidean vacuum for linearized gravity is actually a finite-dimensional algebraic obstruction related to Killing vectors and linearization instabilities, rather than an infinite-dimensional divergence requiring renormalization.
• Mathematical Foundation: By using Calderón projectors, the authors bypass the convergence issues of mode expansions, providing a rigorous construction of the state that is valid on the global Cauchy surface.
• Symmetry Trade-off: The work highlights a fundamental tension in quantum gravity on de Sitter space: one cannot simultaneously have a positive, gauge-invariant, and fully de Sitter-invariant Hadamard state for linearized gravity. The authors show that preserving positivity and gauge invariance necessitates breaking the full de Sitter symmetry (specifically, invariance under time translations/boosts).
• Comparison with Literature: The paper contrasts their "projection" method with previous "mode-expansion" approaches (e.g., Faizal-Higuchi, Allen), showing that while those methods effectively remove the problematic modes by hand, the authors' approach provides a unified geometric framework using spectral theory and Calderón projectors.

Conclusion

Gérard and Wrochna demonstrate that the naive Euclidean vacuum for linearized gravity on de Sitter space is ill-defined due to a finite-dimensional positivity failure linked to Killing 1-forms. They construct a modified, positive, and gauge-invariant Hadamard state by projecting out these modes. This construction is mathematically rigorous but results in a state that is not invariant under the full de Sitter group, resolving a long-standing ambiguity in the quantization of gravity on curved spacetimes.