Positive resolution of Bartnik's cosmological splitting conjecture

Technical Summary: Positive Resolution of Bartnik's Cosmological Splitting Conjecture

Problem Statement
The paper addresses the rigidity aspect of the cosmological Hawking–Penrose singularity theorem. While the singularity theorem (Theorem 1.1) establishes that a cosmological spacetime (globally hyperbolic, with compact Cauchy surfaces, satisfying the strong energy condition $Ric(v,v) \geq 0$) is generically causally geodesically incomplete, the question of what occurs when this incompleteness fails remained open. Specifically, Robert Bartnik (1988) conjectured that if such a spacetime is timelike geodesically complete, it must exhibit maximal rigidity: it must split isometrically as a Lorentzian product $\mathbb{R} \times S$ with metric $dt^2 - \tilde{g}$, where $(S, \tilde{g})$ is a compact Riemannian manifold with nonnegative Ricci curvature. This conjecture posits that the failure of singularity implies the dynamics of the spacetime become "trivial."

Methodology
The authors employ a synthesis of two distinct analytical frameworks to prove the conjecture:

1. Global Viscosity Solutions to the Lorentzian Eikonal Equation:
   The authors utilize the construction by Zhu–Wu–Cui [24] of global viscosity solutions to the Lorentzian eikonal equation $g(\nabla u, \nabla u) = 1$. Instead of basing these solutions on a single timelike line (as in classical splitting theorems), they associate Busemann-type functions $u^+$ and $u^-$ to a foliation of Cauchy surfaces $\{\tau_s\}_{s \in \mathbb{R}}$. These functions are constructed as limits of sequences involving the time separation function $\ell$ between Cauchy surfaces $\tau_{s_j}$ and $\tau_{t_j}$ as $s_j \to -\infty$ and $t_j \to +\infty$.

2. Elliptic Approach via the $p$-d'Alembertian:
   Building on recent joint work with Braun, Gigli, and Sämann [5], the authors apply a degenerate elliptic operator, the $p$-d'Alembertian ($\square_p$), defined for $p < 1$ as $\square_p f := -\text{div}(|\nabla f|_g^{p-2} \nabla f)$. This operator allows the authors to treat the hyperbolic Lorentzian geometry using elliptic techniques (specifically maximum principles) which are not directly applicable to the standard d'Alembertian.

Key Technical Steps and Contributions

• Specialization to Compact Cauchy Surfaces:
  The authors adapt the Zhu–Wu–Cui construction to the setting of compact Cauchy surfaces. By selecting maximizing timelike geodesics between $\tau_{s_j}$ and $\tau_{t_j}$, they identify a sequence of intersection points $z_j$ on a fixed Cauchy surface $\tau_0$. Using compactness, they extract a limit point $z_\infty$ and define the Busemann functions $u^\pm$ relative to this base point.

• Tangency and Non-negativity (Proposition 2.3):
  A critical contribution is the proof that the sum of the forward and backward Busemann functions satisfies $u^+ + u^- \geq 0$ on $M$, with equality holding at the base point $z_\infty$ (i.e., $u^+(z_\infty) + u^-(z_\infty) = 0$). This establishes the necessary "tangency" condition for a maximum principle argument.

• Weak $p$-d'Alembertian Comparison (Proposition 2.4 & 2.5):
  The authors establish a weak comparison principle for the Lorentz distance to a compact spacelike Cauchy surface. They prove that $u^+$ is weakly $p$-superharmonic ($\square_p u^+ \leq 0$) and $-u^-$ is weakly $p$-subharmonic ($\square_p (-u^-) \geq 0$). This relies on the semiconcavity/semiconvexity of the distance functions and the specific properties of the $p$-d'Alembertian for $p < 1$.

• Strong Tangency and Global Equality (Proposition 2.6):
  Combining the tangency condition ($u^+ + u^- \geq 0$ with a zero at $z_\infty$) with the super/sub-harmonicity properties, the authors apply a maximum principle for uniformly elliptic operators (derived from the $p$-d'Alembertian structure). This proves that $u^+ = -u^-$ globally on $M$. Furthermore, this equality implies that the functions are $C^{1,1}_{loc}$.

• Construction of the Timelike Line:
  The equality $u^+ = -u^-$ allows for the concatenation of the calibrated future and past rays associated with $u^+$ and $u^-$ at any point $z$. The authors demonstrate that this concatenation forms a globally maximizing timelike line (a geodesic defined on $\mathbb{R}$ that maximizes the Lorentzian distance between any two of its points).

Main Result
The paper proves Theorem 1.2 (Bartnik's Splitting Conjecture):

If a cosmological spacetime $(M, g)$ is timelike geodesically complete, then it is isometric to a product spacetime $(\mathbb{R} \times S, dt^2 - \tilde{g})$, where $(S, \tilde{g})$ is a compact Riemannian manifold of nonnegative Ricci curvature.

Significance and Claims
The authors claim to provide a complete proof of Bartnik's 1988 conjecture, thereby establishing the rigidity of the cosmological Hawking–Penrose singularity theorem. The significance lies in:

1. Resolving a Long-Standing Open Problem: The conjecture had been verified only under stronger assumptions (e.g., nonnegative timelike sectional curvature) or additional constraints. This work removes those extra assumptions, relying only on the strong energy condition and global hyperbolicity with compact Cauchy surfaces.
2. Methodological Synthesis: The proof successfully bridges the gap between the construction of global viscosity solutions for the Lorentzian eikonal equation (Zhu–Wu–Cui) and the elliptic $p$-d'Alembertian techniques developed in [5]. This combination overcomes the fundamental challenge of the hyperbolic nature of the Lorentzian Laplacian by utilizing the degenerate ellipticity of the $p$-d'Alembertian for $p < 1$.
3. Generalizability: The authors note in the "Outlook" that their methods appear adaptable to weighted Lorentz–Finsler settings and low regularity contexts, suggesting the robustness of the elliptic approach to Lorentzian rigidity problems.

The paper does not claim to solve rigidity for the general (non-cosmological) Hawking–Penrose theorem or to address other causality conditions beyond those specified, maintaining a focused scope on the cosmological splitting scenario.