Conformal boundary rigidity from null geodesic travel times

Technical Summary: Conformal Boundary Rigidity from Null Geodesic Travel Times

Problem Statement
The paper addresses a rigidity question for asymptotically anti-de Sitter (AdS) spacetimes, motivated by the interplay between conformal structure, causal structure, and gravitational mass. Specifically, it investigates whether the causal behavior of null geodesics at conformal infinity uniquely determines the spacetime geometry up to conformal isometry.

The central question (Question 1.2) is formulated as follows: Let $M$ be an asymptotically AdS spacetime with conformal boundary $\mathcal{I}$. If, for every point $p \in \mathcal{I}$, every null geodesic in $M$ departing from $p$ (with an idealized past endpoint at $p$) returns to $\mathcal{I}$ at the unique spacetime antipode $q$ of $p$, is the spacetime conformally isometric to anti-de Sitter space?

This question is framed as a Lorentzian analogue to two classical problems in Riemannian geometry: Blaschke's wiedersehen manifold problem (where all geodesics from a point reconverge at an antipode) and the boundary rigidity problem (determining a metric from boundary distance data). The authors note that positive mass theorems for asymptotically AdS spacetimes often imply rigidity for zero-mass cases; this work seeks to establish such a rigidity result using only causal data (null geodesic travel times) rather than traditional pointwise energy conditions.

Methodology
The authors approach the problem by analyzing the behavior of null geodesics under specific geometric assumptions, progressing from restrictive cases to a general stationary setting.

1. Null Energy Condition (NEC) Approach: In Section 3.1, the authors assume the spacetime admits a representative metric in its conformal class satisfying the Null Energy Condition ($Ric(\ell, \ell) \geq 0$). They utilize the Raychaudhuri equation for null geodesic congruences. By assuming all null geodesics from $p$ reach the antipode $q$ without conjugate points (a consequence of the definition of spacetime antipodes), they demonstrate that the expansion, shear, and rotation must vanish. This forces the Ricci tensor and the Weyl tensor to vanish along all null directions, implying the spacetime is locally conformally flat and Einstein, hence anti-de Sitter.

2. Static Spacetime Approach: In Section 3.2, the authors relax the energy condition but assume the spacetime is static (admitting a hypersurface-orthogonal timelike Killing vector field). They employ the quotient construction to map the spacetime to a Riemannian "optical" or "Fermat" manifold $(\Sigma, \tilde{h})$. Null geodesics in the spacetime project to geodesics in $\Sigma$ with respect to the Fermat metric. The condition that null geodesics travel from $p$ to $q$ with a specific time delay translates to the condition that all geodesics in $\Sigma$ connecting boundary points have a constant length (specifically $\pi$). By invoking a rigidity theorem for Riemannian manifolds with constant geodesic length (Bangert's theorem), they conclude $\Sigma$ is a round hemisphere, implying the spacetime is conformally AdS.

3. Stationary Spacetime Approach (Main Contribution): In Section 4, the authors generalize the static case to globally stationary spacetimes (where the Killing vector field is not necessarily hypersurface-orthogonal).

   • Mapping to Magnetic Geodesics: They map null geodesics in the stationary spacetime to "magnetic geodesics" on the quotient optical manifold. The stationary Killing field introduces a 1-form $\omega$ (related to the shift vector) which acts as a magnetic potential. The travel time of null geodesics corresponds to the Mañé action functional $A(\gamma) = \int (|\dot{\gamma}| - \omega(\dot{\gamma})) dt$.
   • Rigidity of Magnetic Chords: The core technical result (Theorem 4.1) establishes that if every magnetic chord (a magnetic geodesic connecting boundary points) on a compact Riemannian manifold with a totally geodesic round sphere boundary has the same Mañé action ($\pi$), then the magnetic field ($\Omega = d\omega$) must vanish, and the manifold is isometric to a standard round hemisphere.
   • Proof Strategy: The proof involves showing that the scattering relation (mapping entry data to exit data) for these magnetic geodesics must coincide with that of the standard hemisphere. This is achieved by analyzing the first variation of the free-time action, utilizing symplectic geometry (Lagrangian submanifolds in the cotangent bundle), and proving that the magnetic flow is non-trapping under the given action constraints. The vanishing of the magnetic field implies the stationary spacetime is effectively static and conformally AdS.

Key Results

• Theorem 1.3 / Theorem 4.4: The paper provides an affirmative answer to Question 1.2 for globally stationary asymptotically AdS spacetimes. If every null geodesic from a point $p$ on the conformal boundary returns to the spacetime antipode $q$, the spacetime is conformally isometric to anti-de Sitter space.
• Proposition 3.1: The affirmative answer holds if the spacetime satisfies the Null Energy Condition (without assuming stationarity).
• Proposition 3.3: The affirmative answer holds if the spacetime is static.
• Theorem 4.1 (Magnetic Rigidity): A novel result in Riemannian geometry stating that a compact Riemannian manifold with a totally geodesic round sphere boundary, admitting a magnetic field such that all magnetic chords have constant Mañé action, must be a standard round hemisphere with a vanishing magnetic field.

Significance and Claims
The authors claim that this work extends the scope of positive mass rigidity theorems. While traditional positive mass theorems often rely on pointwise energy conditions (like the NEC) to prove that zero-mass spacetimes are rigid (i.e., are exactly AdS), this paper demonstrates that rigidity can be derived purely from causal structure (the behavior of null geodesic travel times) in the stationary case.

The paper emphasizes that the question is motivated by the "holographic causality" property, where the causal structure of the boundary encodes the bulk mass. The result suggests that the "borderline" case between holographic causality and the timelike boundary Penrose property (where null geodesics exactly hit antipodes) uniquely characterizes the vacuum AdS solution.

The authors explicitly note that the question remains open for general asymptotically AdS spacetimes that are neither stationary nor satisfy the NEC. They also discuss an analogous question for asymptotically flat spacetimes (Question 5.2), providing an affirmative answer under the NEC but noting that the stationary/magnetic geodesic techniques do not directly apply due to the non-compactness of the optical manifold in the flat case.

Conclusion
The paper successfully establishes a conformal boundary rigidity theorem for stationary asymptotically AdS spacetimes using a synthesis of Lorentzian geometry, inverse problems, and the theory of magnetic geodesics. It replaces the need for energy conditions with a specific causal condition on null geodesic travel times, reinforcing the link between the causal structure at infinity and the global geometry of the spacetime.